(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Measurement and prediction of herbicide transport into shallow groundwater"

MEASUREMENT AND PREDICTION OF HERBICIDE 
TRANSPORT INTO SHALLOW GROUNDWATER 



By 

MATTHEW CLAY SMITH 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL 
OF THE UNIVERSITY OF FLORIDA IN 
PARTIAL FULFILLMENT OF THE REQUIREMENTS 
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY 



UNIVERSITY OF FLORIDA 



1988 



DEDICATION 



To ray mother, 

Corrine L. Smith 

She was taken from her family by illness during this research. Her 
wisdom, courage, and strength continue to inspire me. Without that inner 
strength, I would never have made it this far. 

I love you Mom. 



AC3<NCWI£DGMENTS 

I would like to express my sincere appreciation to the following: 

Dr. A. B. (Del) Bottcher, chairman of my advisory committee, for 
his friendship, patience, and guidance. He provided valuable advice and 
philosophy during the highs and lows encountered in my Fh.D. program. 

Dr. K. L. Campbell, cochairman of my advisory committee, for his 
guidance and support. His calm and reasoned approach to problem solving 
is a model which I need to emulate. 

Dr. E. D. Threadgill, for serving on my advisory committee, for 
making available the field site on which this research was conducted, 
and for being the best engineer/scientist/adininistrator that I could 
ever hope to be associated with. 

Dr. W. C. Huber, for serving on my advisory committee, and being 
one of the very best instructors that I have encountered during many 
years of college. If all instructors were that good, my GPAs would have 
been higher. 

Dr. P. S. C. Rao, for serving on my advisory committee, for showing 
a sincere interest in my research, and for teaching me several very 
valuable lessons during the qualifying exam. 

Ms. M. W. Smith, for being a wonderful, laving, and supportive 
wife. She assumed a greatly disproportionate share of child care and 
household duties so that I could devote time to this research. 



iii 



Nicholas and Sarah Smith, my children, for their love and for 
providing meaning and purpose to my life and education. 

Dr. G. W. Isaacs, for financial support in the form of a research 
assistantship . 

Dr. D. L. Thomas, for his friendship and support. He shielded me 
from many distractions at work while I completed this document. 

Dr. W. B. Wheeler and Ms. S. J. Scherer, at the UF Pesticide 
Residue laboratory, for allowing me to use their equipment and guiding 
me through the intricacies of pesticide residue analysis. 

Mr. L. A. Asmussen, Dr. R. A. Leonard, Dr. W. G. Knisel, Ms. L. R. 
Marti, and others at the USDA Southeast Watershed Research Laboratory for 
their advice, encouragement, and support. 

USDA, Southern Region Pesticide Impact Assessment Program, for 
financial support in the form of a grant. 

Many others, too numerous to list here, in both Gainesville and 
Tifton who contributed in many ways to the success of the project. 



iv 



TABLE OF CONTENTS 

Page 

ACKNOWUEDGEMENTS iii 

LIST OF TABLES vii 

LIST OF FIGURES viii 

ABSTRACT xiv 

CHAPTERS 

1 rNTRODUCnON 1 

2 OBJECTIVES 4 

3 REVIEW OF THE LITERATURE 5 

3.1 Evidence of Pesticide Residues in Groundwater 5 

3.2 Factors Which Influence Pesticide Transport to 

to Groundwater 7 

3.3 Predicting Pesticide Transport 15 

3.4 Field Studies of Pesticide Transport 27 

3.5 Summary 36 

4 EXPERIMENTAL METHODS 39 

4.1 Field Site Description 39 

4.2 Site Instrumentation 42 

4.3 Chemical Applications 49 

4.4 Sample Collection and Storage 52 

4.5 Sample Analysis 55 

5 MODELING THE EXPERIMENTAL SITE 66 

5.1 Selection of Common Input Parameter Values 67 

5.2 Parameters Unique to PRZM 73 

5.3 Parameters Unique to GLEAMS 77 



v 



6 RESULTS AND DISCUSSION 79 

6.1 Data Collection 79 

6.2 Sample Analysis 83 

6.3 Chemical Applications 85 

6.4 Chemicals in the Unsaturated Zone 91 

6.5 Chemicals in the Saturated Zone 107 

6.6 Model Results and Comparisons 142 

7 SUMMARY AND CONCLUSIONS 157 

8 RECOMMENDATIONS FOR IMPROVEMENTS AND FURTHER STUDY 160 

REFERENCES 162 

APPENDICES 

A. MONITORING WELL STATISTICS 170 

B. HOW TO GET COMPLETE DATA SET 173 

C. WATER BALANCE PROGRAMS 176 

D. SURFACE PLOTS OF ATRAZINE CONCENTRATION IN GROUNDWATER 200 

BIOGRAPHICAL SKETCH 213 



vi 



LIST OF TABLES 

Page 

5.1. Soil properties used in simulations 69 

5.2. Chemical properties used in simulations 70 

5.3. Chemical applications summary 73 

6.1. Chronological summary of field site activities 80 

6.2. Chemical application results 88 

6.3. Measured application rates and soil surface concentrations 

of atrazine and alachlor 91 

6.4. Mean concentrations of atrazine (mg/kg) in soil samples 96 

6.5. Mean concentrations of alachlor (mg/kg) in soil samples 97 

6.6. Mean concentration of bromide (mg/L) at each sampling 
following first application 105 

6.7. Mean concentration of bromide (mg/L) at each sampling 
following second application 106 

6.8. Total mass of chemicals in the saturated zone 133 

6.9. Simulated mass flux of chemicals 142 

A. l. Monitoring well statistics 171 

B. l. Example listing of the water sample data set 174 

B. 2. Example listing of the soil sample data set 175 

C. l. Sample output from program ANALYZE 191 

C.2. Sample output from program FLUX 198 

vii 



LIST OF FIGURES 

Page 

4.1. Contour maps of soil surface and restricting layer 

showing locations and ID labels of monitoring wells 41 

4.2. Cross-section of soil profile through application area 
showing locations of monitoring wells and soil solution 
samplers 45 

6.1. Uniformity of atrazine application 86 

6.2. Uniformity of alachlor application 87 

6.3. Uniformity of bromide application on 4/27/87 89 

6.4. Comparison of alachlor and atrazine concentrations in 
application samples 90 

6.5. Correlation between atrazine concentrations in the top 5 cm 

of soil and application solution concentrations 93 

6.6. Correlation between atrazine concentrations in the top 5 cm 

of soil and application rate 93 

6.7. Correlation between alachlor concentrations in the top 5 cm 

of soil and application solution concentrations 94 

6.8. Correlation between alachlor concentrations in the top 5 cm 

of soil and application rate 94 

6.9. Bromide and atrazine concentrations in solution sampler 
09N-2 as a function of total water applied since 
application 98 

6.10. Bromide concentration for six sampling locations following 

the first application at a 61 cm depth 101 



viii 



6.11. Bromide concentration for six sampling locations following 

the first application at a 122 cm depth 101 

6.12. Bromide concentration for six sampling locations following 

the first application at a 183 cm depth 102 

6.13. Average bromide concentration for the three sampling depths 
following the first application 102 

6.14. Bromide concentration for six sampling locations following 

the second application at a 61 cm depth 103 

6.15. Bromide concentration for six sampling locations following 

the second application at a 122 cm depth 103 

6.16. Bromide concentration for six sampling locations following 

the second application at a 183 cm depth 104 

6.17. Average bromide concentration for the three sampling depths 
following the second application 104 

6.18. Bromide concentration in the groundwater on 12/15/86 108 

6.19. Bromide concentration in the groundwater on 12/22/86 108 

6.20. Bromide concentration in the groundwater on 12/29/86 109 

6.21. Bromide concentration in the groundwater on 01/05/87 109 

6.22. Bromide concentration in the groundwater on 1/12/87 110 

6.23. Bromide concentration in the groundwater on 1/19/87 110 

6.24. Bromide concentration in the groundwater on 5/05/87 Ill 

6.25. Bromide concentration in the groundwater on 5/08/87 Ill 

6.26. Bromide concentration in the groundwater on 5/13/87 112 

6.27. Bromide concentration in the groundwater on 5/18/87 112 

6.28. Bromide concentration in the groundwater on 5/25/87 113 

6.29. Bromide concentration in the groundwater on 6/01/87 113 

ix 



6.30. Atrazine concentration in the groundwater on 2/23/87 115 

6.31. Atrazine concentration in the groundwater on 3/02/87 115 

6.32. Atrazine concentration in the groundwater on 3/09/87 116 

6.33. Atrazine concentration in the groundwater on 3/16/87 116 

6.34. Atrazine concentration in the groundwater on 3/23/87 117 

6.35. Atrazine concentration in the groundwater on 3/31/87 117 

6.36. Contour plot of water table elevation on 5/08/87 showing 
direction of flow 118 

6.37. Concentration of atrazine and water table elevation in 

well 09-11 120 

6.38. Nitrate concentration in groundwater on 5/01/87 122 

6.39. Nitrate concentration in groundwater on 5/03/87 122 

6.40. Nitrate concentration in groundwater on 5/05/87 123 

6.41. Nitrate concentration in groundwater on 5/08/87 123 

6.42. Nitrate concentration in groundwater on 5/13/87 124 

6.43. Nitrate concentration in groundwater on 5/18/87 124 

6.44. Nitrate concentration in groundwater on 5/25/87 125 

6.45. Nitrate concentration in groundwater on 6/01/87 125 

6.46. Chloride concentration in groundwater on 5/01/87 127 

6.47. Chloride concentration in groundwater on 5/03/87 127 

6.48. Chloride concentration in groundwater on 5/05/87 128 

6.49. Chloride concentration in groundwater on 5/08/87 128 

6.50. Chloride concentration in groundwater on 5/13/87 129 

6.51. Chloride concentration in groundwater on 5/18/87 129 

6.52. Chloride concentration in groundwater on 5/25/87 130 

6.53. Chloride concentration in groundwater on 6/01/87 130 



x 



6.54. Total mass of atrazine stored in the saturated zone 133 

6.55. Subareas used in water balance 135 

6.56. Comparison of percolation volumes predicted by GLEAMS and 

PRZM 144 

6.57. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 11/18/86 145 

6.58. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 11/24/86 145 

6.59. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 12/22/86 146 

6.60. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 2/09/87 146 

6.61. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 3/16/87 147 

6.62. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 5/25/87 147 

6.63. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 11/18/86 148 

6.64. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 11/24/86 148 

6.65. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 12/22/86 149 

6.66. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 2/09/87 149 

6.67. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 3/16/87 150 



xi 



6.68. Comparison of measured and FRZM predicted alachlor 
concentrations in the soil on 5/25/87 150 

6.69. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 61 cm depth following the first 

application 152 

6.70. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 122 cm depth following the first 

application 152 

6.71. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 183 cm depth following the first 

application 153 

6.72. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 61 cm depth following the second 

application 153 

6.73. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 122 cm depth following the second 

application 154 

6.74. Measured and FRZM predicted bromide concentrations in the 
soil solution at a 183 cm depth following the second 

application 154 

6.75. Measured and FRZM predicted concentrations of atrazine in 

the soil solution at a 61 cm depth 155 

C.l. Listing of program to calculate water and chemical 

fluxes and storages 177 

C. 2. Program to calculate mass balance between sampling periods.. 194 

D. l. Atrazine concentration in the groundwater on 1/19/87 201 



xii 



D.2. Atrazine concentration in the groundwater on 1/26/87 201 

D.3. Atrazine concentration in the groundwater on 2/02/87 202 

D.4. Atrazine concentration in the groundwater on 2/09/87 202 

D.5. Atrazine concentration in the groundwater on 2/16/87 203 

D.6. Atrazine c»noentration in the groundwater on 2/23/87 203 

D.7. Atrazine concentration in the groundwater on 3/02/87 204 

D.8. Atrazine concentration in the groundwater on 3/09/87 204 

D.9. Atrazine concentration in the groundwater on 3/16/87 205 

D.10. Atrazine concentration in the groundwater on 3/23/87 205 

D.ll. Atrazine concentration in the groundwater on 3/31/87 206 

D.12. Atrazine concentration in the groundwater on 4/06/87 206 

D.13. Atrazine concentration in the groundwater on 4/13/87 207 

D.14. Atrazine concentration in the groundwater on 4/20/87 207 

D.15. Atrazine concentration in the groundwater on 4/30/87 208 

D.16. Atrazine concentration in the groundwater on 5/01/87 208 

D.17. Atrazine concentration in the groundwater on 5/03/87 209 

D.18. Atrazine concentration in the groundwater on 5/05/87 209 

D.19. Atrazine concentration in the groundwater on 5/08/87 210 

D.20. Atrazine concentration in the groundwater on 5/13/87 210 

D.21. Atrazine concentration in the groundwater on 5/18/87 211 

D.22. Atrazine concentration in the groundwater on 5/25/87 211 

D.23. Atrazine concentration in the groundwater on 6/01/87 212 



xiii 



Abstract of Dissertation Presented to the Graduate School 
of the University of Florida in Partial Fulfillment of the 
Requirements for the Degree of Doctor of Philosophy 

MEASUREMENT AND PREDICTION OF HERBICIDE 
TRANSPORT INTO SHAUOW GROUNDWATER 

By 

Matthew Clay Smith 
December 1988 

Chairman: Dr. A. B. Bottcher 

Major Department: Agricultural Engineering 

A field study was conducted to observe the movement of the 
herbicides atrazine and alachlor within the soil profile and a shallow 
water table aquifer following a surface application. Surface-applied 
bromide was used as a nonadsorbed tracer of water movement. The movement 
of nitrate from a fertilizer application to the site was also monitored. 
Measurements of chemical concentration were made using soil core samples, 
samples of soil water from the unsaturated zone, and water samples from 
monitoring wells below the water table. 

Atrazine was observed to move rapidly with both saturated and 
unsaturated flows. Concentrations of atrazine exceeded 350 jxg/L in soil 
water samples at a depth of 61 cm. Samples of shallow groundwater 
contained atrazine residues as high as 90 jizg/L. Measurable 
concentrations of alachlor did not move below a depth of 45 cm in the 
soil. 



xiv 



Bromide and nitrate concentrations in soil water demonstrated large 
variability between sampling locations. Nitrate concentrations in the 
groundwater exceeded 40 mg/L after fertilization of the field. 

Two root/vadose zone pesticide transport models, GLEAMS and 
ERZM, were used to simulate the conditions on the site. Comparisons were 
made between model simulation results and observed data. The 
uncalibrated models predicted concentrations that were generally within 
an order of magnitude, and often were within a factor of 2 to 3, of 
observed values. Differences between the predictions of the models 
appear to be due to the relative detail by which each model describes 
soil properties as a function of depth. 



xv 



CHAPTER 1 
rNTRODUCnCN 



Agricultural production systems have undergone many changes since 
the 1940s. The capability of post-war industry to synthesize almost 
unlimited varieties of chemicals in great quantities has provided 
farmers with an arsenal with which to battle insects, weeds, and 
disease. These chemicals can also provide plants with required 
nutrients and regulate their growth. The result of using these chemical 
tools along with improved machines for planting, cultivating, and 
harvesting has been a dramatic increase in the productivity of the 
American farmer. 

However, some serious problems associated with the use of agricul- 
tural chemicals have been identified in recent years. Many of these 
chemicals are toxic to non-target organisms such as birds, fish, and the 
person who is applying them. Rachel Carson (1962) brought to the 
attention of the public the potential harmful effects of using these 
chemicals in large quantities over large land areas. Since the 
publication of her book 'Silent Spring 1 there has been much research 
conducted to evaluate the fate of agricultural chemicals in the 
environment. EXiring the 1960s and 1970s, great effort was initiated to 
determine the quantities of soil, fertilizers and pesticides which were 
entering our lakes and rivers. Through these efforts methods have been 



1 



2 

developed (though only a few have been implemented) to minimize or 
eliminate agricultural impacts on surface waters. 

The discovery of abandoned waste sites such as the famous Love 
Canal and the discovery of the pesticide aldicarb in the shallow 
groundwater on Long Island in New York generated concern for the 
protection of groundwater. These concerns led to analyzing samples of 
groundwater for toxic wastes and pesticides. 

As more groundwater samples were analyzed, it became apparent that 
agricultural chemicals had somehow entered groundwater aquifers in 
several areas of the country. These results were publicized, and 
citizens throughout the nation expressed alarm that they were drinking 
potentially toxic chemicals. The federal and state governments 
responded to these concerns by intensifying monitoring efforts and 
reviewing data on the many agricultural chemicals to determine potential 
for leaching to groundwater and the extent of the current problem. 
Several federal agencies, along with university and private industry 
scientists, began developing computer models to help explain why these 
chemicals were moving beyond their target locations and to screen 
chemicals for their relative mobilities in soil. A limiting factor in 
the development and utilization of these models was a lack of the 
detailed data needed to test and validate model predictions. 

To date, data on the presence of agricultural chemicals in 
groundwater have been limited primarily to sampling of municipal and 
domestic wells, providing only a large scale picture of contamination. 
There have been only a few intensive monitoring studies completed which 
provide sufficient data for model parameter estimation. Currently, there 



are increased efforts at specific sites to provide more detailed 
monitoring of chemical movement through the soil profile and within 
groundwater aquifers. 

The study described here is an effort to develop data for use in 
the development and testing of models which describe the fate and 
transport of pesticides and nutrients used in agriculture. 



CHAPTER 2 
OBJECTIVES 

The overall objective of this research is to develop data on the 
movement of two widely used herbicides through the crop root zone, 
through the vadose zone (unsaturated zone between the crop root zone and 
the top of the unconf ined water table aquifer) , and within the shallow 
water table aquifer. These data will then be compared with the 
predictions of two root/vadose zone pesticide transport models. 

The specific objectives are to perform the following: 

1. Instrument a field site for monitoring the movement of water and 
chemicals through the soil profile and a shallow, water table 
aquifer. 

2. Observe the fate and transport of surface-applied herbicides 
(atrazine and alachlor) over time. 

3. Use surface-applied bromide as a nonadsorbed tracer of water 
movement. 

4. Observe transport of the nitrate component of fertilizers 
applied to field. 

5. Compile the observations of chemical movement into a data base 
for use in testing pesticide transport models. 

6. Simulate the field conditions present during the monitored 
period with two pesticide transport models and compare 

to observed results. 



4 



CHAPTER 3 
REVIEW OF THE LITERATURE 

3.1 Evidence of Pesticide Residues in Groundwater 

The U. S. Environmental Protection Agency estimates that at least 19 

pesticides have been detected in groundwater in 24 states as a result of 

agricultural practices (U.S. EPA, 1987a) . The total number of 

pesticides detected in groundwater is greater than nineteen. These 

additional detections are the result of manufacturing, storage, and 

loading activities which are not included in the above figures. The 

number of reported cases of pesticides in groundwater is increasing. 

Cohen et al. (1986) suggest that the increase is due to improvements in 

the quality and quantity of studies instead of an increase in the 

problem. 

The insecticide/nematicide aldicarb was detected in the shallow 
groundwater on Long Island, New York in 1979 (Wartenburg, 1988) . 
Aldicarb was used as an insecticide on potatoes. Aldicarb was detected 
in 20 of 31 wells tested in 1978 and the presence was confirmed through a 
repeat sampling of the wells in June, 1979. According to the U.S. EPA 
(1987a) , aldicarb residues have been detected in almost 2000 wells on 
Long Island. Concentrations of aldicarb exceeded the state health 
guideline of 7 /xg/L (ppb) in nearly 50% of those wells. Concentrations 
as high as 515 yjg/L were recorded in the Long Island wells as reported by 
Ritter (1986) . Aldicarb residues have been found in groundwater in 15 
states (Cohen et al., 1986) with typical concentrations reported in the 
range of 0.3 - 3 jug/L. 

5 



6 

In the state of Florida, residues of the nematicide EDB have been 
detected in over 10% of the public and private drinking water wells which 
have been sampled (U.S. EPA, 1987a). Approximately 1,200 wells require 
treatment or have been closed as sources of drinking water. Marti et al. 
(1984) detected EE© in the groundwater in southwest Georgia. EDB 
residues have been detected in groundwater in at least 6 other states. 

In California, residues of approximately 57 pesticides have been 
detected in groundwater (U.S. EPA, 1987a) . Most of these detections 
were related to factors other than routine use on agricultural fields. 
The nematicide DBCP has been detected at levels exceeding 1 jug/L in 
approximately 60% of the 2500 drinking water wells tested. 

Residues of the widely used herbicide atrazine are commonly 
identified in groundwater. Ritter (1986) reports that approximately 
36,000,000 kg of atrazine were used in the United States in 1982. 
Atrazine has been detected in PA, IA, NE, WI, MN, and MD (U.S. EPA, 
1987; Cohen et al., 1986; Ritter, 1986) . According to Berteau and Spath 
(1986) , atrazine has also been detected in concentrations of up to 2 /xg/L 
in California groundwater. Pionke et al. (1988) reported atrazine 
concentrations in groundwater at levels up to 1.1 /xg/L in an agricultural 
watershed in Pennsylvania. The wells sampled by Pionke et al. (1988) 
were all located in unconf ined aquifers with depths to the water table 
ranging from 2 to 21 m. Atrazine concentrations of 10 /xg/L have been 
reported in a karst aquifer in Iowa (Libra et al., 1986; Kelley et al., 
1986) . 

Residues of the herbicide alachlor have been found in groundwater in 
the states listed for atrazine, excluding WI. According to Ritter 



7 

(1986) , alachlor and atrazine together accounted for 25% of all 
pesticides sold in the U. S. in 1982. Ritter (1986) reported that in 
1982, approximately 38,600,000 log of alachlor were applied to cropland in 
the United States. Kelley et al. (1986) reported alachlor concentrations 
in groundwater as high as 16 /xg/L in Iowa. 

There are many individual cases of pesticide residue detection in 
groundwater that could be cited. The reader is referred to the 
proceedings of the Agricultural Impacts on Groundwater Conferences 
(National Water Well Assoc. , 1986, 1988) for a variety of reports 
concerning the detection of pesticide residues in groundwater. A book 
edited by Garner et al. (1986) is also a good source of information 
related to this subject. 

3.2 Factors Which Influence Pesticide Transport To Groundwater 
Donigian and Rao (1986a) list five processes which affect the fate 
and movement of chemicals within the soil. These processes are 
transport, sorption, transformation/degradation, volatilization, and 
plant uptake. Other authors, e.g. , Cheng and Kbskinen (1986) , use 
slightly varied groupings which include the processes named above. The 
interactions of these processes over time and space determine the fate of 
chemicals in the soil. The spatial variability of soil properties and 
other factors is an important consideration when interpreting field data 
and model predictions (Donigian and Rao, 1986a) . This section will 
describe each of the processes and same of the factors which influence or 
regulate them. 



8 

3.2.1 Transport 

The chemicals can be transported in at least three different phases: 
adsorbed to solid materials, in solution, and as a vapor. Surface runoff 
can transport chemicals in both the adsorbed and solution phases. 
Erosion and transport of soil particles during rainfall and irrigation 
can result in significant transport of highly-adsorbed, low-solubility 
pesticides (Wauchope, 1978) . For most pesticides, the majority of 
transport in runoff is in the solution phase (Wauchope, 1978; Rao and 
Davidson, 1980) . Chemicals in the solution phase can be transported 
within the soil profile by saturated and unsaturated water flows. 

Water percolating through the soil profile can transport chemicals 
in the solution phase beneath the root zone and possibly into 
groundwater. The most common methods used to describe the transport of 
water and solutes in porous media are based on the assumptions that 
Darcian flow conditions exist, and that solute movement is controlled by 
advective and dispersive processes (Freeze and Cherry, 1979; Jury, 1986a; 
Wagenet, 1986; Wang and Anderson, 1982). Darcian flow conditions imply 
that the flow equations are representative of some volume of soil. All 
soil properties such as pore size distribution and volumetric water 
content are assumed to be uniform throughout the volume of soil (Jury, 
1986b) . 

The advective component of transport is usually based upon Richards' 
equation and describes the average rate of water flow. The average rate 
of solute flow is the product of the solute concentration and the average 
pore-water velocity. Additional transport of solute can occur due to 
mechanical mixing of water in adjoining pores during advective transport 



9 

and molecular diffusion of solute from pores with high concentrations to 
adjoining pores with lower concentrations (Freeze and Cherry, 1979) . 
Molecular diffusion can be significant when average pore-water 
velocities are low; otherwise, the processes involved in mechanical 
dispersion usually dominate. Differences in pore-water velocities in the 
direction of bulk flow cause a spreading out, and consequently a lowering 
of peak concentrations, of the solute plume. Some of the solute will 
arrive at a reference point earlier, and some will arrive later, than 
would be predicted based upon the average linear flow velocity. This is 
referred to as longitudinal dispersion. The tortuosity and 
interconnection of pores in the soil will also cause the solute to move 
perpendicularly to the bulk flow direction. This is referred to as 
transverse dispersion. Freeze and Cherry (1979) and Rao et al. (1988) 
provide detailed discussions of the dispersive-diffusive processes and 
their mathematical formulations. 

Transport predictions based upon the methods described above have 
not compared well with field observations, e.g. Jury et al. (1986) . 
Numerous modifications to the general theory have been proposed in order 
to help explain and reduce the differences between field observations and 
theoretical predictions. Rao et al. (1980) proposed separating pore- 
water in soils into two regions: inter-aggregate and intra-aggregate. 
Convective-dispersive transport was assumed to be limited to the water in 
the inter-aggregate region. The intra-aggregate region was assumed to 
act as a source/sink of solute. Diffusion between the regions could 
remove or add solute to the water in the inter-aggregate region. Jury 
(1982) proposed a method of predicting the transport times of solutes to 



10 

specified depths which is not based on physical processes. Rather, the 
transfer function theory uses measured values to develop a probability 
density function of travel times and uses it to predict solute travel 
times to depths greater than those that were measured. Gish (1987) 
combined the convective-diffusive transport theory with a stochastic 
representation of water velocities to describe field-scale (average) 
concentrations of bromide at various depths in the soil profile. 
Comparisons between the average measured bromide concentrations and 
predicted values showed good agreement. 

Volatilization from soil and plant surfaces transports the chemicals 
into the atmosphere and reduces the amount remaining for transport with 
runoff or percolation. Diffusion of chemicals in the vapor phase can 
also transport the chemical within the soil profile. Most transport 
models do not consider vapor diffusion. 

Many of the factors which are assumed to influence the transport of 
chemicals through porous media are summarized by Jury (1986b) . Some of 
the soil properties influencing transport are water content, bulk 
density, permeability, clay content, organic matter content, and water 
retention (field capacity) . Helling and Gish (1986) present results of a 
simple modeling exercise to demonstrate the effect of several soil 
properties on pesticide transport. Some of the environmental factors 
which have been shown to influence transport are precipitation, 
evapotranspiration, and temperature (Jury, 1986b) . 
3.2.2 Sorption 

The partitioning of solutes between the liquid and solid phases 
(dissolved and adsorbed) is a major factor determining the mass of 



11 

solute available for advective-dispersive transport through the soil 
profile. If the chemical of interest is strongly adsorbed, then only a 
small fraction will exist in the solution phase at a given time and be 
available for transport. Non-adsorbed chemicals, however, will exist 
entirely in the solution phase and are available for transport with 
percolating water. 

Many pesticides are nonpolar and adsorption occurs primarily on 
organic matter surfaces. Pesticides and chemicals which are polar will 
adsorb primarily to clay surfaces (Jury, 1986c) . Since the majority of 
pesticides are nonpolar, the adsorption onto soils has been primarily 
related to the organic matter or organic carbon content of the soil. A 
detailed discussion of the factors influencing pesticide adsorption onto 
soils is given by Jury (1986c) . The relationship between concentrations 
of a pesticide in the dissolved and sorted phases is often represented by 
the Freundlich equation: 

S = HC" 3.1 
where S = adsorbed concentration (/xg/g of soil) , C = solution 
concentration (Mg/mL) , and K and n are empirical constants for the soil- 
pesticide system. The exponent n in equation 3.1 is often assumed to be 
equal to 1.0 which results in a linear relationship of the form: 

S - KdC 3.2 
where = partition coefficient (ml/g of soil) . 

For most uses of equation 3.2 in transport modeling, the 
partitioning between the solution and adsorbed phases is assumed to be 
instantaneous and reversible. Errors associated with the assumptions of 



12 

linearity, instantaneous equilibrium, and reversibility are discussed by 
Rao and Davidson (1980) . 

The partition coefficient, is unique to a given pesticide-soil 
combination. However, Rao and Davidson (1980) report that investigators 
have shown that when is normalized for the organic carbon content of 
the soil, the resulting value is independent of the soil type and can be 
considered a property of the pesticide. The normalized partition 
coefficient is represented as and is defined by: 

Kqc = Kd * 100 / %OC 3.3 
where %OC is the percent organic carbon content of the soil and is the 
measured partition coefficient. Methods for measurement of the 
partition coefficient are reviewed by Rao and Davidson (1980) . 

3.2.3 Volatilization 

Volatilization of pesticides from the soil and plant surfaces 
reduces the amount of the pesticide available to be leached or 
transported with runoff. Volatilization also determines the quantity of 
the pesticide which exists in the vapor phase within the soil and thus 
the amount available for diffusive vapor transport. Thus, as noted by 
Donigian and Rao (1986a) , volatilization affects both the fate and 
transport of a pesticide. 

Some of the factors which influence volatilization are summarized by 
Jury and Valentine (1986) . They are Henry's constant %, chemical 
concentration, adsorption site density, temperature, water content, wind 
speed, and water evaporation. Henry's constant, % is the ratio of 
saturated vapor density to solubility and is an index of the partitioning 



13 

between the vapor and solution phases. A larger implies increased 
volatilization. Increasing concentrations of the pesticide will increase 
the volatilization as long as the vapor density is not saturated. 
Adsorption of the pesticide will reduce volatilization. Volatilization 
increases as temperature increases. As the soil water content decreases, 
the rate of volatilization increases. Increasing wind speed can increase 
the volatilization of pesticides, particularly those with low % values. 
Evaporation of water from the soil surface can transport pesticides from 
within the soil profile to the surface where temperature and wind 
effects can increase the volatilization. Each of these factors is 
discussed in more detail by Jury (1986d) . 

3.2.4 Transformation/degradation 

Transformation of a pesticide is the change in structure or 
composition of the original compound and degradation is the breakdown of 
the compound into smaller fragments with eventual inorganic endproducts 
such as H 2 and C0 2 (Cheng and Kbskinen, 1986) . The transformation and 
degradation processes represent a loss of the original (or parent) 
compound, thus reducing the amount remaining in the soil for transport by 
surface runoff and percolation. 

In general, transformation and degradation processes occur at faster 
rates on plant surfaces and in the top few centimeters of soil than in 
the deeper soil zones (Donigian and Rao, 1986a) . Thus, a chemical that 
is foliar or surface-applied will likely degrade more rapidly than if it 
were incorporated. For non-persistent pesticides (half-lives less than 
15-20 days) , the timing of rainfall events or irrigation is important in 



14 

determining the fraction of the applied mass that will be available for 
transport. Events occurring shortly after application will likely result 
in the largest concentrations in runoff and percolation water (Wauchope, 
1979; Donigian and Rao, 1986a) . The timing of rainfall and irrigation 
events is less critical for persistent (half-lives in excess of 100 days) 
pesticides since they will reside in the soil for longer periods 
(Donigian and Rao, 1986a) . 

The major processes involved in transformation and degradation 
include biotransformation, chemical hydrolysis, photolysis, and 
oxidation-reduction (Donigian and Rao, 1986a) . Factors influencing 
transformation and degradation of pesticides, as summarized by Jury and 
Valentine (1986) , include: microbial populations, chemical concentration, 
temperature, oxygen, pH, soil water content, and light. Detailed 
descriptions of the many factors influencing these processes are 
presented by Rao and Davidson (1980) , Valentine and Schnoor (1986) , and 
Valentine (1986) . 

3.2.5 Plant processes 

Plant processes relating to pesticides are very complex. The 
uptake, translocation, acxwmulation, and transformation of pesticides by 
plants affect the availability of pesticides for transport processes and 
the potential exposure to pesticide residues by the consumers of the 
vegetation, fruits, etc. (Donigian and Rao, 1986a) . Plant processes can 
serve as both a sink and a source of pesticide residues available for 
transport. 



15 

Pesticides applied to plant foliage will likely be transformed or 
degraded more rapidly than if the pesticide were applied to the soil or 
incorporated. Pesticides on the foliage my also be absorbed into the 
plant. Pesticides in the solution phase may be taken up by plant roots 
and translocated to various parts of the plant. These processes serve as 
a loss mechanism reducing the availability of the pesticide for 
transport. 

Pesticide residues on plant foliage can be dislodged by rainfall and 
irrigation and washed onto the soil surface, resulting in an increased 
mass of pesticide available for transport in runoff and percolation 
(Smith et al. , 1981; Donigian and Rao, 1986a) . Plant residues which are 
left on the soil surface or incorporated can also serve as a source of 
pesticide which may be available for transport. 

3.3 Predicting Pesticide Transport 
There are many methods that can be used to predict the mobility of a 
given pesticide. These range from simple indices to complex research 
models. Each method has value when used for the purpose for which it was 
designed and with recognition of the assumptions and limitations 
associated with it. 

3.3.1 Indices of contamination potential 

Rao et al. (1985) compared several methods for computing indices of 
the contamination potential of pesticides. Most of the indices are based 
on one or more chemical properties such as and solubility. Three 
indices also included distance to groundwater and recharge rate. The 



16 

attenuation factor (AF) proposed by Rao et al. (1985) can also be used to 
estimate the mass of pesticide which will leach below the root zone or 
out of the vadose zone and into groundwater. The AF index is the only 
one of the indices compared by Rao et al. (1985) which incorporated mass 
loadings. 

Dean et al. (1984) developed a methodology in which soil 
characteristics, crop type, chemical properties, and management practices 
(e.g. tillage type) can be combined to find a cumulative frequency 
distribution of the percentage of applied pesticide mass leaching below 
the crop root zone. These frequency distributions were derived from 
hundreds of 25-year simulations of pesticide leaching using the RRZM 
model (Carsel et al., 1984) . The methodology was applied to four crop 
types (corn, soybean, cotton, and wheat) and 19 representative growing 
regions. The results are somewhat unique in that the user is presented 
with a statistical probability of leaching based on a 25-year 
meteorologic record. 

Another index which is currently receiving much attention is the 
DRASTIC index developed by Aller et al. (1985) . This index does not 
consider chemical properties; instead, it is an index of the 
vulnerability of groundwater at a given location to contamination from 
surface-applied chemicals. The DRASTIC index assigns numerical scores or 
weights to seven factors which could influence pollution potential. The 
factors are depth to water, net recharge, aquifer media, soil media, 
topography, impact of vadose zone, and conductivity of the aquifer. The 
final DRASTIC score is used to describe an area as having high, medium, 
or low susceptibility to groundwater pollution. 



17 

ERASTTC has been used to map every county in the United States 
(Alexander and Little, 1986) as part of the first stage of a national 
survey of pesticide contamination of drinking water (U.S. EPA, 1987b) . 
In the first county-level assessment the Southeast Coastal Plain, which 
covers all of Florida and southern Georgia, was shown to be the 
groundwater region of highest vulnerability. 

Indices like those mentioned above are beneficial to regulatory 
agencies as a method to screen great numbers of compounds and locate 
vulnerable areas. This information can aid in allocating funding for 
further studies concentrating on the chemicals of interest in regions 
which may be most vulnerable. These studies may involve detailed 
modeling of pesticide transport to groundwater, modeling of surface and 
groundwater hydrology, and collection of data from test locations. 

3.3.2 Transport models 

There are many models of pesticide transport reported in the 
literature. These range from steady-state screening models, e.g. PESTAN 
(Enfield et al. , 1982) , to very complex process-oriented research models, 
e.g. LEACHMP (Wagenet and Hutson, 1986) . Donigian and Rao (1986a) , and 
Shoemaker and Magette (1987) review a number of the models which are 
available. Most models which are capable of simulating pesticide fate 
and transport do not explicitly represent agricultural management and 
cropping practice effects on runoff and leaching. Modifications of 
management practices will be one method of reducing pesticide leaching to 
groundwater. The models used for regulatory purposes should reflect the 
trade-offs between surface water and groundwater quality. There is some 



18 

evidence that no-till farming will increase pesticide transport to 
groundwater while reducing surface runoff and thereby reduce pollution 
of surface waters (Dick et al. , 1986). Thus, solutions to one problem 
may exacerbate another problem. The models discussed below are capable 
of simulating, to some degree, agricultural management practice effects 
on pesticide fate and transport. 

According to Donigian and Rao (1986a) , SESOIL (The Seasonal Soil 
Compartment model) is designed for long-term simulation of pesticide fate 
in the soil environment. It was developed for the EPA and is used as a 
screening model. SESOIL can simulate many components of the hydrologic 
cycle including precipitation, evapotranspiration (ET) and surface 
runoff. The model considers transport within the unsaturated zone 
extending from the soil surface to the top of the saturated zone. The 
hydrologic responses are determined using physically based equations in 
which uncertainty has been included. The water balance used in the 
model is a statistical representation of the hydrologic components over a 
"season." A season is the time step of the model, e.g., month or year. 
Erosion is simulated using the Universal Soil Loss Equation (Wischmeier 
and Smith, 1978) . Hetrick and Travis (1988) coupled SESOIL with EROS (a 
submodel of the CREAMS (Khisel, 1980) watershed model) to predict surface 
runoff and sediment yield from small watersheds. The pesticide component 
of the model considers most of the processes and factors influencing 
pesticide transport which were described in previous sections. 

SESOIL requires calibration and represents long-term averages. Same 
of the results of use and testing of SESOIL reported by Donigian and Rao 
(1986a) indicate that it should not be used for short-term predictions. 



19 

The combined SESOIL/EROS model of Hetrick and Travis (1988) adequately 
predicted long-term (several months) average runoff and sediment yields 
when tested against data from three small watersheds. However, monthly 
predictions were not good, especially when most of the runoff occurred 
due to large single storm events. 

MOUSE (Method Of Underground Solute Evaluation) was developed at 
Cornell University as a management model and a training tool for students 
and professionals (Steenhuis et al . , 1987). MOUSE is an interactive, 
menu-driven program which can run on a I EM-PC or compatible computer. 
The model can read daily historical weather files or generate synthetic 
rainfall, air, and soil temperature patterns. Surface runoff is computed 
using the SCS curve number method. Erosion is not considered. Solute 
movement in the vadose zone considers advective and dispersive flux, as 
well as degradation and adsorption. MOUSE will also simulate the 
movement of a pesticide in a two-dimensional unconfined aquifer. 
Extensive graphics illustrate movement as it is being simulated. 

PRZM (Pesticide Root Zone Model) is a field-scale hydrology and 
transport model developed by the EPA (Carsel et al., 1984) . PRZM is a 
continuous model capable of simulating water and chemical fluxes over 
many years of historical daily weather records. Runoff is predicted 
based on the SCS curve number equation, and erosion is simulated using a 
modification of the Universal Soil Loss Equation for daily time steps. 
The model can simulate the entire vadose zone (soil surface to 
groundwater) . The vadose zone can be characterized by several layers 
with varying properties. For calculations, the vadose zone is divided 
into many compartments of equal depth. The model simulates crop growth 



20 

(leaf area and rooting depth) . Water in the root zone can be removed by 
percolation, evaporation, or transpiration by the plant. Percolation is 
calculated based on the water-holding capacity of the soil. When the 
water content in a layer exceeds field capacity (1/10 - 1/3 bar tension) 
the excess water drains into the next lower compartment. There is an 
option in PRZM that allows the draining of the profile to occur over a 
longer period than one day. 

Pesticide processes represented include advective and dispersive 
flux, sorption, degradation in soil and on plant foliage, and plant 
uptake. Volatilization and transport in the vapor phase are not 
considered. Application of pesticides can be partitioned between the 
soil surface and plant foliage. Applications to the soil can be 
incorporated by tillage. Different degradation rates can be specified 
for soil and foliar pesticide residues. The degradation rate within the 
soil can also be varied by soil layer. 

PRZM will simulate multiple applications of one pesticide each year 
for many years of continuous climatic record. Thus it can predict 
temporal variations in leaching and runoff. Effects of agricultural 
management and cropping practices can be simulated. PRZM allows the user 
to request time series results for many variables in the model. This 
feature is useful for observing processes at intermediate locations 
within the vadose zone over time. 

GLEAMS (Groundwater Loading Effects of Agricultural Management 
Systems) is also a field-scale hydrology and transport model (Leonard et 
al. , 1987) . It is based upon the extensively documented and applied 
CREAMS model (Knisel, 1980) . CREAMS is a nonpoint source model for 



21 

predicting sediment, nutrient, and pesticide losses with surface runoff 
from agricultural management systems. GLEAMS builds upon the foundations 
in CREAMS by adding components to simulate movement of water and 
chemicals within the crop root zone. like CREAMS, GLEAMS is a 
continuous, daily simulation model. 

GLEAMS predicts runoff using the SCS curve number method. Erosion 
is predicted based upon modifications to the Universal Soil Loss 
Equation. Surface runoff and eroded sediment, with chemicals in both the 
dissolved and sorted phases, can be routed overland, in channels, and 
through impoundments. 

GLEAMS divides the crop root zone into seven layers. The vadose 
zone between the root zone and the water table is not considered. The 
first layer has a thickness of 1 cm. This layer is assumed to be the 
portion of the soil that determines the mass of pesticide available for 
extraction into surface runoff. The thickness of the second layer is 
1/6 of the root zone depth minus 1 cm for the first layer. The remaining 
5 layers are each 1/6 of the root zone in thickness. This layering 
structure is fixed by the program. A strong correlation has been 
observed between pesticide concentrations in runoff and the concentration 
of pesticides in the top 1 cm of the soil profile (Leonard, 1988) . Thus 
the authors of CREAMS and GLEAMS believe that a 1 cm active zone at the 
soil surface is required to maintain sensitivity of runoff concentrations 
to soil concentrations (Leonard and Khisel, 1987) . Soil profile 
properties can vary with depth. The model weights the input values to 
establish average properties for each layer in the model. Percolation 
through the profile is based on the water-holding capacity of the soil as 



22 

in PRZM. Water in excess of field capacity drains to the next lower 
layer. 

Pesticide transport within the root zone is by advection. No 
dispersive flux components are included. Volatilization is not 
considered. Pesticide applications can be partitioned between plant 
foliage and the soil surface. Different degradation rates can be 
specified for chemicals on the foliage and within the soil. Soil-applied 
chemicals can be incorporated to a specified depth. 

The model will simulate up to ten chemicals simultaneously. This 
feature makes possible the observation of the effect of changes in 
chemical properties, e.g. partition coefficient, on resulting leaching 
losses with one model run. The model can also simulate the formation, 
fate, and transport of degradation products of the parent chemicals. 

3.3.3 Coupled saturated/unsaturated zone transport models 

The root/vadose zone models described above are important tools for 
assessing the mobility of pesticides within the unsaturated zone. The 
concern expressed by the public and governmental officials relate to 
pesticide residues in groundwater supplies used for drinking water. Thus 
the unsaturated zone models need to be linked to saturated zone transport 
models so that the concentrations and transport of pesticides within 
aquifers can be calculated. 

As described above, M3USE (Steenhuis et al. , 1987) can predict the 
transport of pesticides in a vertical, two-dimensional cross section of 
a water table aquifer. Jones et al. (1987) coupled the PRZM model to a 
two-dimensional saturated zone transport model to predict the movement of 



23 

aldicarb residues within a shallow water table aquifer. Dean and Carsel 
(1988) reported on progress in linking FRZM to a two-dimensional 
saturated transport model. The linked model will use FRZM for simulating 
the root zone, a one-dimensional transport model for the vadose zone, and 
a two-dimensional saturated zone model which can simulate confined, 
unconfined, and leaky confined aquifers. The saturated zone model will 
also simulate pumping from the aquifer (s) . The linked model is expected 
to be released for testing during the fall of 1988. 

3.3.4 Model testing and uncertainty 

Methods (such as described above) to predict the transport of the 
chemicals used by agriculture are needed by regulators in order to assess 
the impacts on surface water and groundwater quality from the use of 
these chemicals. Field testing of the contamination potential of all 
chemicals registered for use by agriculture would be infeasible due to 
both cost and time constraints. Since decisions regarding chemical use 
will be made based upon the results of transport models, it is important 
to understand the errors and uncertainty associated with the models. 
Each model or method will have errors associated with the assumptions 
relating to the description of the system, choice of equations to 
represent the system, and input parameters required by the models. There 
are same general points that can be made regarding the uncertainties 
inherent in deterministic, physically based models such as FRZM and 
GIEAMS which were described previously. An excellent discussion of the 
uncertainty in transport models is presented by Leonard and Knisel 
(1988). 



24 

Limitations of the conceptual formulation of the system that is 
described by a model lead to uncertainty in the results. Most 
deterministic models assume that some of the physical properties of the 
system are uniform with respect to location. That is, soil properties 
such as porosity, water-holding capacity, and organic matter content are 
constant across the field, although most models will allow these types of 
properties to vary with depth in the soil profile. Furthermore, 
variations in pore sizes within the soil, which can cause uneven water 
velocity distributions, are usually not considered. Each model will 
represent a system (ie. the crop root zone) in varying levels of detail. 

The equations which are used to represent the fate and transport 
processes have limitations and uncertainties. For example, PRZM and 
GLEAMS both use the SCS curve number method for partitioning rainfall 
between runoff and infiltration. This method utilizes daily rainfall 
records and does not consider the effects of rainfall intensity on the 
runoff process. Testing of the CREAMS model (Khisel, 1980) , from which 
GLEAMS was derived, showed that the daily hydrology option represented 
average annual runoff volumes well, but did not do as well with daily and 
monthly runoff volumes (Smith and Williams, 1980) . FRZM and GLEAMS both 
utilize modified versions of the USLE (Wischmeier and Smith, 1978) for 
prediction of daily soil erosion. The USLE was developed for long-term 
(20-yr) average annual erosion rates. Although these modifications have 
been tested, the basis of these methods is the USLE which is an empirical 
equation derived for long-term average predictions. Many transport 
models utilize the linear form of the Freundlich equation (Equation 3.1) 
to describe the partitioning of a chemical between the dissolved and 



25 

adsorbed phases. Rao and Davidson (1980) indicate that the assumption of 
linearity can lead to errors on the order of a factor of two to three. 
The previous examples illustrate that the mathematical representations of 
processes used in models often have inherent inaccuracies and 
uncertainties. 

The parameters which are required by the models to solve the 
equations which describe the system also have inherent uncertainties 
associated with them. Jury, (1985) , in a review of the literature, 
reported coefficients of variation (CV) from field measurements of 
several soil physical properties. Soil porosity was found to have a mean 
CV of 10%. Bulk density measurement had a CV of 9%. Determinations of 
particle size fractions had an average CV of 28%, with a range of 3 to 
55%. The water-holding capacity of the soil at matric potentials of 0.1 
and 15 bars had a CV of 15 and 25%, respectively. Measurements of 
saturated hydraulic conductivity and infiltration rates showed a CV of 
124 and 72%, respectively. In a study by Jury et al. (1986), the 
partition coefficient, K^, of napropamide measured on 36 samples from a 
1.44-ha field had a CV of 31%. Rao and Davidson (1980) reported that the 
normalized partition coefficient, K^, reported from individual studies 
generally had a CV of 40-60%. The variability of the properties reported 
above are not necessarily due to the fact the samples were taken over 
large areas. Jury (1985) reported that one author had reported that up 
to 50% of the variation in a parameter over a field occurred within a 1 
m 2 area. 

Given the uncertainties associated with models and the parameters 
required by the models as outlined briefly above, how closely must a 



26 

model match field observations in order to be considered validated 
(determined to accurately represent the system)? Leonard and Knisel 
(1988) indicate that there are no standard criteria for model validation. 
Since the spatial distribution of variations in soil properties is 
usually unknown, it would be unreasonable to expect any model to exactly 
match point measurements from a field study. Hedden (1986) reported that 
participants of the Predictive Exposure Assessment Workshop sponsored by 
the U. S. EPA in Atlanta, GA, on April 27-29, 1982, agreed on two 
criteria for model acceptance. For screening applications of the model 
(limited site-specific data and not calibrated to previous data from the 
site) , the model should be able to replicate measured field data within 
an order of magnitude. For site-specific applications (parameters 
measured on-site and with the model calibrated to the site) , the model 
should be able to match the field observations within a factor of two. 
The screening level criteria seems to be quite reasonable when all of the 
sources of error and uncertainty are considered. The site-specific 
criteria, however, may be difficult to meet by even the best models using 
carefully measured site-specific parameters. 

The fact that models can not be expected to exactly replicate field 
measured values does not mean that they have no value. Models that have 
been verified (ie. their response to changes in parameters reflect what 
would be expected based on knowledge of the system) can be used with some 
confidence to evaluate the differences in predicted leaching of different 
chemicals or, perhaps, differences in leaching of a single chemical due 
to different climatic or management conditions. 



27 

3.4 Field Studies of Pesticide Transport 
Field-scale studies of pesticide runoff and leaching have, until 
recently, been very limited. Studies of sufficient detail to be used for 
the validation and testing of transport models have been limited due to 
three factors, according to Donigian and Rao (1986b) : 

1. Most field studies have focused on either chemical leaching or 
surface runoff from watersheds; few studies have examined both. 

2. Since sampling and analysis costs associated with pesticide 
leaching are high, most studies are of short duration. 

3. Most studies were designed for purposes other than developing 
data sets for model testing and therefore are incomplete with 
regard to measurement of all the required model input parameters. 

In this section, four field studies are reviewed. Methods of 
sampling and monitoring will be emphasized. Same of the studies were 
conducted specifically for model validation and testing purposes. A 
review of the Big Spring Basin study is included due to the unique 
conditions in the study area. 

3.4.1 Aldicarb in Florida citrus groves 

A three-year field study of the movement and degradation of 
aldicarb in both the unsaturated zone and a shallow water table was 
conducted at two locations in Florida (Hornsby et al., 1983; Jones et 
al., 1987) . The two locations are referred to as the Oviedo site and the 
lake Hamilton site. At both locations, aldicarb was applied to bedded 
citrus trees located on coarse-textured soils. Some of the soils at the 
Oviedo site had a thick organic layer overlying the coarse textured 



28 

subsoil. The Oviedo site is in the flatwoods area and is poorly drained. 
The Lake Hamilton site was located on a sand ridge. Drainage of the 
upper soil layers was very rapid. The Oviedo site encompassed 3.6 ha of 
treated area and the other site encompassed approximately 1.7 ha. 

Aldicarb was banded along each side of the trees and incorporated to 
a depth of 5 cm. At the Oviedo site, soil samples were collected in 30 
cm increments to a depth of 150 cm. At the Lake Hamilton site, soil 
samples were collected in 30 cm increments to a depth of 60 cm, and then 
in 60 cm increments to a depth of 300 cm. Soil samples were collected 
using a bucket auger and were transferred into plastic bags for storage 
at -20°C until analyzed. The soil samples were collected monthly for six 
months following the first application of aldicarb in 1983. Undisturbed 
soil cores and bulk soil samples were also collected from untreated areas 
at each site for laboratory determination of soil and chemical 
properties. 

At each site, two clusters of observation wells were installed 
within the treated area, and additional well clusters were located 
upgradient and downgradient of the treated area. The wells were screened 
at depths of 2, 3, 5, and 6 m at the Oviedo site, and at depths ranging 
from 5 to 15 m, depending on location, at the Lake Hamilton site. Water 
samples were usually collected with a peristaltic pump. At least 5 well 
volumes were pumped prior to collecting a sample for analysis. When 
flowrates into the well were too slow to permit pumping of 5 well 
volumes, the wells were evacuated and the water entering the wells was 
pumped for an additional 5 minutes prior to collection of the sample. 
The groundwater temperature, pH, and conductivity were measured during 



29 

sampling. Water samples were collected monthly throughout the study. 
The storage conditions prior to analysis were not specified. 

Results from the first 6 months of the study were reported by 
Hornsby et al. (1983) . Results are reported in terms of total toxic 
residues (TTR) , which is the sum of residues of aldicarb, aldicarb 
sulfoxide, and aldicarb sulfone. Fifteen days after application, TTR 
were not detected below 30 cm at the Oviedo site, but were detected in 
all replicates at the lake Hamilton site to a depth of 120 cm. After 120 
days, TTR were detected at the deepest sampling depths (150 cm at Oviedo, 
and 300 cm at Lake Hamilton) at both sites. No TTR were observed at the 
Oviedo site in any of the observation wells during the first six months. 
Two wells within the treated area at the Lake Hamilton site showed 
measurable levels of TTR. In one well, TTR levels as high as 1.2 mg/L 
were measured approximately 130 days after application. Data relating to 
rainfall or irrigation during the study were not presented. 

Jones et al. (1983) used the data from the aldicarb study above to 
evaluate three pesticide transport models. One of the models tested was 
FRZM (Carsel et al., 1984) . Agreement between FRZM predicted movement of 
aldicarb and the observed data was good. A sensitivity analysis was 
performed to demonstrate the response of FRZM to variations in several 
soil and pesticide properties. Jones et al. (1987) presented results 
from the Lake Hamilton site in which water samples from the shallow water 
table were collected from a network of 174 wells over a period of three 
years. These data were used to evaluate predictions from a linkage of 
the FRZM model to a one- or twc»-dimensional saturated transport model. 
The linked model was developed to allow predictions of the extent of 



30 

lateral transport of TTR which have entered a shallow water table. PRZM 
was used to predict daily mass loadings of aldicarb to the top of the 
water table. Results of the saturated zone modeling were in reasonably 
good agreement with the observed concentrations in the wells and the rate 
of lateral movement. 

3.4.2 Comparison of tillage effects in Maryland 

Paired watersheds were established in 1984 on the Eastern Shore of 
Maryland to observe the impacts of tillage practices on agricultural 
chemical leaching and runoff. The experimental design and early results 
have been reported by Brinsfield et al. (1987, 1988) . One of the 
objectives of the experiments was to generate a data set for testing and 
validation of pesticide transport models. One watershed was used to grow 
corn using conventional tillage (CT) practices, and on the other 
watershed corn was grown using no-till (NT) methods. The CT watershed is 
approximately 6 ha in size, and the NT watershed covers approximately 8.9 
ha. The soils in both watersheds are silty, well-drained, and nearly 
level. 

Soil samples were collected at 5 locations in each watershed 4 times 
per year. The samples were collected using a hand auger in 15 cm 
increments to a depth of 120 m. 

Monitoring wells made of 6.35 cm diameter PVC were installed to 
depths of 3-4 m on 30 m centers in each field. The wells were arranged 
to divide each field into four quadrants. The wells were installed so 
that only the top of the water table would be sampled. A bentonite-soil 
mixture was used to seal the top 0.5 m of each well. Well samples were 



31 

collected monthly. Hie wells were pumped for 5 minutes prior to 
collecting a sample for analysis. 

To monitor surface runoff, H-flumes with automatic samplers and 
recorders were installed in the outlets of the watersheds. All samples 
were frozen for transport and storage prior to analysis. Details of the 
analytical procedures used for pesticide residue analysis are presented 
by Brinsfield et al. (1987) . 

Gravity-fed lysimeters were installed horizontally through the walls 
of a reinforced pit to collect samples of percolating water. The 
lysimeters were installed at a depth of 1 m below the soil surface. 
Three lysimeters were installed in each pit. These lysimeters collect 
samples of saturated flow percolating through the root zone. Samples 
were collected from these lysimeters after each rainfall event. 

Herbicide treatments were the same for both watersheds. Atrazine, 
metolachlor, and simazine were applied pre-emergence at the rate of 1.68 
kg AI (active ingredient) /ha. Cyanazine was applied at 2.24 kg Al/ha, 
and carbofuran at 1.12 kg Al/ha, both pre-emergence. Dicamba was 
applied post-emergence at a rate of 0.55 kg Al/ha. 

Leachate collected in the lysimeters during January, 1985, had 
atrazine concentrations as high as 2 jug/L. After planting in 1985, no 
leachate samples were collected in the CT watershed until September. 
Atrazine and metolachlor concentrations in those samples ranged from 1-2 

Simazine was detected in only one sample, and cyanazine was not 
detected. In the NT watershed, samples collected in July had 
concentrations of atrazine at 8-10 /ig/L, simazine at 7-10 jig/L, cyanazine 
at less than 1 /ig/L, and metolachlor at less than 2 /ig/L. Leachate from 



32 

the NT watershed in the fall of 1985 had similar c»noentrations to those 
collected from the CT watershed. 

Groundwater samples collected 15 days after application in 1984 
showed levels of atrazine, simazine, and cyanazine exceeding 1 jug/L. The 
concentration of atrazine in one well reached 7 ^ig/L on this date. 
IXaring late summer and fall, atrazine was the only pesticide detected in 
the wells. Pesticide concentrations in the groundwater were similar for 
the two watersheds, although the concentrations were somewhat higher 
during the growing season in the NT watershed. In April, 1986, (prior to 
application) atrazine concentrations of 4 jug/L were observed in the water 
table on both watersheds. This indicates that atrazine may be persistent 
in the lower soil zones and in the water table. 

3.4.3 Aldicarb movement in the Douahertv Plain of Georg ia 

The U. S. Environmental Protection Agency and the U. S. Geological 
Survey began a study in 1982 to observe the movement of aldicarb and its 
residues in the unsaturated and saturated zones at a field site in Lee 
County, in southwestern Georgia. Aldicarb is used as an insecticide and 
nematicide on peanuts grown on the site. One of the primary objectives 
was to develop an extensive database of soil, land use, geologic, and 
pesticide data for use in the validation and testing of the PRZM (Carsel 
et al., 1984) model. Different aspects of this study have been reported 
by a number of investigators (e.g. Cooper, 1986; Rao et al., 1986; Hook, 
1987; Hedden, 1986; Smith and Carsel, 1986). 

The study site covers an area of approximately 4.5 ha. Four soils 
are mapped in the field. All have a fine-loamy texture. Below a depth 



33 

of approximately 1 m there are layers of clay separating zones of coarser 
sand and gravel. The conductivity of the clay layers is very low, which 
suggests that perched water tables may form on these layers and cause 
lateral transport of leached pesticides. There is a shallow water table 
at the site which can vary in position by as much a 6 m during a year. 

Soil samples were collected from the site to characterize soil 
properties and pesticide sorption and decay characteristics. Geologic 
information was collected from holes bored to the top of the Ocala 
Limestone. The residuum was approximately 13 m thick. The Ocala 
Limestone is a part of the Floridan aquifer system. Four wells were 
cased into the Floridan aquifer for measurement of water levels. 
A weather station was installed at the site to collect meteorological 
data. 

Statistical procedures were used to determine that 20 monitoring 
sites would accurately reflect the fate and transport of aldicarb at this 
site. The locations of the monitoring sites were randomly selected and 
distributed between the three major soil series based upon the relative 
area in each series. 

Each monitoring site was equipped with an impressive array of 
instruments for monitoring of conditions and collecting samples. Five 
tensiometers were installed to a depth of 150 cm for monitoring soil 
water content. The tensiometers were monitored and serviced three times 
per week. Five thermistors were installed to a depth of 115 cm for 
measurement of soil temperatures. The thermistors were monitored three 
times per week during the middle of the afternoon in order to estimate 
maximum soil temperatures. Three soil solution samplers were installed 



34 

at depths of 1.5, 2.1, and 2.7 m. Hie samplers were made using stainless 
steel for the body, a high-flow ceramic cup, and teflon tubing for sample 
collection and applying the vacuum. A silica flour slurry was poured 
around each sampler to insure good contact with the surrounding soil. 
The vacuum and sample lines were buried to avoid interference during 
field operations. Vacuum was applied to the samplers for a period of 24 
hours prior to collection of the sample. 

At 15 of the monitoring sites, permanent stainless steel monitoring 
wells were drilled to a depth of approximately 4.6 m. A stainless steel 
screen 0.6 m in length was placed at the bottom of each well. A gravel 
pack was placed around the screen and a cement grout was placed from the 
gravel to the top of the finished well. The wells were finished so that 
the top of the well was approximately 0.6 m below the soil surface so as 
not to interfere with cropping operations. After field operations are 
complete, the wells are extended above the soil surface. Water samples 
are collected using a dedicated teflon tube (one per well) and a 
peristaltic pump. The 250 mL glass sample containers were connected in 
line between the well and the pump to prevent cross contamination. The 
wells were typically pumped for 5-7 minutes prior to collecting a sample. 

Soil samples were collected using a hand auger in 15 cm increments 
to a depth of 120 cm. After each sample was collected, gravel and plant 
material was removed and then the sample was sealed in a metal container. 

The references which were obtained for this review did not report 
any results of pesticide monitoring from the field site. The detailed 
monitoring network and experimental design used at this site should 
produce a valuable data set for model validation and testing. 



35 



3.4.4 Big Spring Basin in Iowa. 

The Big Spring Basin is a 267 km 2 groundwater basin located in 
northeastern Iowa. The basin is entirely agricultural with 
approximately 60% of the land area in row crops. The basin has been 
extensively monitored and characterized, and results from these studies 
have been presented by a number of investigators (Libra et al . , 1986; 
Kelley et al., 1986; Hallberg, 1986; Libra et al., 1987). Over 85% of 
the basin's groundwater is discharged through Big Spring. Thus, 
investigators have an easily accessible point for monitoring average 
groundwater quality in the basin as affected by agricultural activities. 
Big Spring is a karst spring with an estimated 10% of flow contributed by 
directed recharge through sinkholes. By using hydrograph separation 
techniques, the investigators can determine the water quality effects of 
both normal infiltration and direct recharge through the sinkholes. 

Four years of monitoring have shown flow-weighted mean 
concentrations of N0 3 -N ranging from 7-11 mg/L. Combined losses of N0 3 -N 
in both surface water and groundwater amount to 33-55% of the average 
annual fertilizer nitrogen applied in the basin. Historic records 
indicate that the magnitude of N0 3 -N concentrations have increased by 
200-300% over the last 20 years. This corresponds to a 200-300% increase 
in the application of nitrogen fertilizers during this period. 

The herbicide atrazine is the only pesticide that is always present 
in the discharge from Big Spring. Flow-weighted mean concentrations of 
atrazine are less than 1 /xg/L, but the average concentrations have 
steadily increased during four years of monitoring. Total losses of 



36 

atrazine in the groundwater amount to less than 0.1% of the annual 
application. Pesticide concentrations in surface waters are generally an 
order of magnitude higher than observed in the discharge from Big Spring. 
Peak concentrations of pesticides occur following recharge events in the 
spring after crops have been planted. 

3.5 Summary 

There is a growing body of evidence that pesticide residues are 
being leached to groundwater in some areas of the country. The U. S. EPA 
recognizes that at least 19 pesticides have been found in groundwater in 
24 states as a result of agricultural practices (U. S. EPA, 1987a) . Many 
of the pesticide detections in groundwater have been from the use of 
nematicides, which in general need to be very soluble in order to be 
effective. The other major class of compounds found in groundwater from 
agricultural practices is herbicides. The herbicides atrazine and 
alachlor are used over large areas in many regions of the country, and 
have been frequently detected in groundwater. Nitrates from 
agriculturally applied fertilizers have also been found in groundwater. 
The health risks associated with the low concentrations (typically less 
than 10 Atg/L) of pesticides in drinking water are still unknown. There 
are documented health risks, however, to infants drinking water with high 
levels of nitrates. The agricultural production systems in this country 
will continue to require large inputs of chemical fertilizers and 
pesticides. Scientists and governmental regulators must identify ways to 
protect groundwater supplies from contamination while allowing farmers to 
use the chemical inputs required to maintain yields. Simulation models 



37 

will play an important role in evaluating the contamination potential 
from applications of pesticides and fertilizers and assess the benefits 
of modifications to current agricultural management practices 
(application methods, timing, chemical formulations, tillage practices, 
etc.) . 

There are many factors which influence the mobility and persistence 
of pesticides in the environment. Soil physical properties, chemical 
properties, crop characteristics, environmental factors, and the 
interactions among them determine the fate of agriculturally applied 
chemicals. Many of the processes involved in the transport and 
degradation of pesticides in the environment are poorly understood, and 
the mathematical equations used to represent these processes are 
simplifications based on experimental observations. The soil properties 
and environmental factors which strongly influence the fate of a chemical 
can vary significantly within small areas. 

Methods to predict the fate of agriculturally applied chemicals 
range from simple indices to complex mathematical research models. 
Models which can represent the effects of agricultural management 
practices on the transport of chemicals are being used by governmental 
regulators as part of the chemical registration process. Most of the 
management models developed to date are designed to represent the 
processes on the soil and plant surfaces, within the crop root zone and 
possibly the unsaturated zone between the root zone and the water table. 
New model developments will incorporate groundwater transport models to 
assess how far and at what rate a chemical that leaches into groundwater 
will move. There are many sources of uncertainty and error associated 



38 

with model predictions. Thus models should not be expected to provide 
exact predictions of transport and fate within a field. There have been 
a limited number of studies however, which provide data of sufficient 
detail to assess how well the models do represent average transport 
within a field. In order to gain confidence in the use of models for 
pesticide fate and transport predictions, many new studies in various 
regions throughout the country are needed. To be of real value in model 
testing and validation, these studies should last for several years. For 
models which require calibration, a year or more of data may be needed to 
calibrate the models and then several additional years of data from that 
site will be needed to test the predictions of the calibrated models. 



CHAPTER 4 
EXPERIMENTAL METHODS 

4.1 Field Site Description 

The study site is located near Tifton, Georgia, in the coastal plain 

physiographic region of the southeastern U.S. The geology of this area 

has been described by Asmussen et al. (1986) . The site is part of a farm 

rented by the University of Georgia Coastal Plain Experiment Station 

which is referred to as Gopher Ridge in honor of the many gopher 

tortoises which make their home on the edges of the fields and in the 

woods nearby. 

The study site covers 0.7 ha in the northwest corner of a 2.3-ha 
field. The field was developed for irrigated agricultural research. 
Irrigation sprinkler risers are located on a 12 by 12 m grid. Supply 
lines for irrigation water are buried approximately 1 m below the soil 
surface. The field was removed from active crop research in 1982. At 
that time the field was deep tilled in both north-south and east-west 
directions to remove possible residual tillage effects. A permanent 
bahia grass cover was established and fertilized and irrigated as needed. 
No pesticides were applied after the grass cover was established. The 
field was maintained in this condition until the spring of 1986 when 
instrumentation for this research was installed. The site is bounded on 
the north and west by trees. Just beyond the tree line on the western 
edge of the field, the surface elevation drops rapidly into a seepage 



39 



40 

area where subsurface flows reemerge as surface water that flows to the 
Little River. 

The soil on the study site is classified as a lakeland sand (Typic 
Quartz ipsamments , thermic, coated) . The soil profile depth on this site 
ranges from 1.9 to 4.4 meters. Underlying the soil is a restricting 
layer consisting of tight clays. This restricting layer is the top of 
the Hawthorne formation which forms the confining layer over much of the 
aquifer system (Floridan) of southern Georgia and Florida. The presence 
of this restricting layer causes percolating rainfall or irrigation to 
saturate the soil above the layer and form a transient water table. 

Previous work by Asmussen et al. (1986) had shown, through the use 
of ground penetrating radar (GPR) , that the restricting layer below the 
study site would cause saturated flow above it to converge into channels. 
The presence of these channels made the site look promising for saturated 
zone monitoring because samples from wells located in these channels 
would be very likely to show the presence of any chemicals which were 
moving with the saturated flow. The maps drawn from the (GPR) 
descriptions also indicated that the slope of the restricting layer near 
the study area boundaries was generally away from the study area. This 
would minimize influences of saturated flows coming into the site from 
other areas. 

Figure 4.1 shows the elevations of both the soil surface and 
restricting layer for the 0.7-ha study site. The elevations given are 
relative to a local bench mark which was assigned an elevation of 30.48 
m. The soil surface has a relatively uniform slope of approximately 4% 
towards the west. The restricting layer shows a more complicated 



41 



Soil Surface Elevation (m) 




Distance (m) 



Restricting Layer Elevation (m) 




Distance (m) 



Figure 4.1. Contour maps of soil surface and restricting layer 
showing locations and ID labels of monitoring wells. 



42 

topography. Slopes of the restricting layer range from 1 to 15 percent, 
with the general direction of the slope also being towards the west. The 
fact that the soil is a coarse sand also was considered to be 
advantageous for monitoring pesticide movement. The high hydraulic 
conductivities and low clay and organic matter contents should result in 
increased leaching and movement of applied chemicals as compared with 
tighter soils. Another advantage of this site was the availability of 
irrigation which could be used to supplement the natural rainfall if 
required to maintain a dynamic flow regime. 

This study was not intended to represent typical agricultural 
chemical usage and practices in the coastal plain of Georgia. The 
primary objective was to observe chemical movement within the unsaturated 
soil and in a shallow water table aquifer (shallow groundwater) . The 
properties of the soil and restricting layer described above were suited 
for this objective. 

4.2 Site Instrumentation 
The movement of the applied chemicals was monitored using a 
combination of soil samples, soil-water (soil solution) samples from the 
unsaturated zone, and saturated zone samples. These samples were 
collected using augers, soil solution samplers, and monitoring wells, 
respectively. Soil sample collection required no permanent 
instrumentation, and the collection procedure will be described in a 
later section. 

Samples of the water from unsaturated soil were collected using 
soil solution samplers. An excellent review of soil solution samplers is 



43 

presented by Litaor (1988) . Smith and Carsel (1988) presented a design 
for a stainless steel solution sampler that do not react with pesticides 
during sampling. Samplers constructed of stainless steel were not 
economically feasible in this research. 

The samplers used in this study were constructed of 4.2 cm OD, 
schedule 40 PVC and a 1 bar, high flow ceramic cup, 3.99 cm diameter by 
19.05 cm long attached to the PVC pipe with epoxy. A 0.64 cm 
polypropylene tube was extended from the inside bottom of the ceramic cup 
to a bulkhead fitting on the top of the sampler. A second fitting was 
placed on the side of sampler, near the top, for attaching a vacuum line. 
The samplers were constructed in two lengths, 1.08 and 2.09 m. The 
polypropylene tubing used in the samplers was not tested for adsorption 
of atrazine and alachlor. This was not done since the solution samplers 
were to be used for collection of samples to be analyzed for inorganic 
chemicals such as bromide and nitrate. The volumes of samples collected 
from the solution samples were in general too small (less than 50 mL) to 
permit reliable quantification of pesticide residues. 

The samplers were placed such that the center of the ceramic cup 
was located at depths of 61, 122, and 183 cm. The samplers were 
installed by auger ing a 5 cm hole to a depth that was approximately 10 
cm less than the desired sampling depth. A piece of thin wall aluminum 
tubing with ID slightly smaller than the ceramic cup was then driven 20 
cm beyond the augered hole, removing a core of soil. The sampler was 
pushed into the hole until seated on the bottom. A slurry of water and 
soil taken from a borrow pit adjacent to the site was then poured into 
the annulus between the PVC and soil. The top 12 cm of the hole was 



44 

widened to approximately 10 can in diameter and was filled with a 
bentonite slurry to form a seal around the sampler and prevent direct 
flow from the surface down the side of the FVC pipe to the ceramic cup. 

Figure 4.1 shows the location of the application area, which is the 
strip to which the herbicides and bromide tracer were applied. The 
samplers and three monitoring wells (08-09, 09-09, and 10-09) are within 
this area. The samplers were installed in groups of three at a distance 
of 3.05 m on either side of each well along a north-south axis. The 
samplers were assigned identification labels that indicated their 
position with respect to the wells and the depth to which they were 
installed, eg. the group of samplers located 3.05 m south of well 09-09 
were assigned labels of 09S-2, 09S-4, and 09S-6 for depths of 61, 122, 
and 183 cm (2, 4, and 6 ft) , respectively. A cross-section of the 
application area showing the locations of solution samplers and 
monitoring wells is shown in Figure 4.2. 

A small instrument trailer was located approximately 4 m east of the 
application area. A vacuum pump located in the trailer was used to 
apply a vacuum to all samplers simultaneously through a manifold. The 
pump was capable of creating a vacuum of approximately 28 cm of Hg. A 
0.64 cm diameter polypropylene sample tube was connected to the sample 
fitting on each sampler and routed back to the trailer where it was 
connected to a 50 mL polycarbonate sample container. All sample tubing 
was enclosed in a protective FVC conduit which was located 1.5 m from the 
samplers to avoid interference with chemical application in the vicinity 
of the samplers. Between the samplers and the FVC conduit, the stiff 
polypropylene vacuum and sample lines were fastened together and held 



45 




Figure 4.2. Cross-section of soil profile through application area 
showing locations of monitoring wells and soil solution 
samplers. 



about 0.6 m above the soil surface to minimize interference around the 
samplers. When it was time to collect the samples from the samplers, 
the vacuum was switched to the sample collection side and the vacuum side 
was opened to atmospheric pressure. Any water that had accumulated in 
the samplers would then be pulled into the sample containers. 

Monitoring wells for collection of samples from the saturated zone 
and measurement of the water table elevation were installed during May 
of 1986. The wells were constructed of 6.3 cm diameter, schedule 40 FVC 
and were slotted using a veneer blade on a radial arm saw. Three rows of 
slots were cut along the circumference of the well. Each slot was 
approximately one-sixth as long as the pipe circumference. The slots 
were spaced approximately 2.5 cm apart over the bottom 1.22 m of the 
well. Although several manufacturers offer slotted FVC well screens, 



46 

these were beyond the budgetary constraints of this project. Some 
existing wells, installed by the USDA-AR3 Southeast Watershed Research 
Laboratory in conjunction with the GPR mapping by Asmussen et al. (1986) 
were also used. Conversations with the technicians involved with the 
installation revealed that the wells were constructed of the same 
materials as described above but were perforated with tiny drill holes 
instead of being slotted. They were unable to recall the diameter of the 
holes or the extent of the perforation above the bottom of the well. 
Appendix A lists statistics associated with each well and indicates which 
wells were installed by USEA. 

The wells were installed by hand auger ing a 10 cm diameter hole 
down to the top of the restricting layer. The restricting layer was 
identified by a sudden change from yellowish-white sand to red clay 
mixed with small rocks. The change was very abrupt and the effort 
required to auger through the clay was significantly greater than to 
auger through the coarse sand. A PVC end cap was slipped onto the end 
of the well and held in place by friction during installation of the 
well. Fine gravel was placed around the wells in the zone of the slots 
in an attempt to minimize sand migration into the wells through the 
relatively coarse slots. Soil from a nearby borrow pit was used to fill 
the hole around the well to within 12 cm of the surface. Ten centimeters 
of a water and bentonite slurry was then added to form a seal around the 
well. The final 2 cm was filled with soil. The USDA augered down to 
the restricting layer with a hollow stem auger on a small drill rig. 
The well was slipped into the center of the auger and the auger was 



47 

withdrawn. No details on backfill procedures used by the USDA are 
available. 

The wells were continuous from the restricting layer to a height of 
approximately 50 cm above the soil surface. No solvent weld joints or 
connections were used. Each well was fitted with a 0.64 cm diameter 
polypropylene tube which extended from approximately 2.5 cm above the 
bottom of the well through a #2 rubber stopper placed in a vented EVC 
cap. The polypropylene tubes were not tested for adsorption of atrazine 
and alachlor. The caps were vented by milling small channels into 
opposite sides of the inside of the cap. This was done to prevent a 
vacuum from forming in the well during sampling. 

The wells were installed on a 12 by 12 m grid corresponding to the 
locations of existing sprinkler risers. The wells were offset by 
approximately 50 cm to the west of the sprinkler risers to avoid hitting 
supply lines beneath the soil surface. In the application area the 
wells and solution samplers were offset from the sprinkler risers by 
about 3.05 m so that the risers and supply lines would not interfere 
with the downward movement of chemicals within the vicinity of the wells 
and samplers. The monitoring wells were assigned ID labels to indicate 
their position relative to the sprinkler riser in the SE corner of the 
2.7 ha field on which the experimental site is located. Figure 4.1 shows 
the locations of the wells and the associated ID labels. Well 08-15 is 
located at the edge of the tree line bordering this side of the study 
site. One well is located in the woods approximately 30 m west northwest 
of well 08-15. This area is a seepage zone where subsurface flows 



48 

reemerge as surface water and flow to the Little River. This well is 
referred to as "LCW". 

Additional instrumentation installed at the site consists of a 
weighing rain gage which was located near well 09-12, a water table 
recorder which was located near well 08-12, tensiometers for measuring 
the water content of the unsaturated soil, and a "deep" well located 
between wells 08-13 and 09-13. 

The rain gage is a US Weather Bureau standard weighing bucket rain 
gage with a seven day chart and 30.5 cm capacity. The rain gage was 
calibrated using standard calibration weights after installation. 

The water table recorder was installed to provide a continuous 
record of the water table elevation to supplement the weekly 
measurements taken on all wells. The recorder was equipped with a seven 
day clock and a drum type chart recorder. 

Eight tensiometers were installed around each set of solution 
samplers. The tensiometers were water filled and connected with very 
small tubing to a mercury manometer board. The tensiometers were 
installed at depths of 30, 60, 90, 110, 122, 140, and 183 cm. There were 
two tensiometers located at the 60 cm depth. The tubing connecting the 
tensiometers to the manometer boards was routed above the soil surface 
with the vacuum and sample lines as noted above. 

The "deep" well was installed by drilling a 10 cm diameter hole 
inside of a 11.4 cm diameter PVC casing. The casing was pushed down as 
the auger advanced. The casing was installed in this manner using 1.5 m 
sections to a depth of approximately 12.2 m below soil surface. The 
last 3 m of the well was located in a saturated formation that liquified 



49 

during drilling. The observation well was made of a 1.5 m section of 
6.3 cm diameter commercial PVC well screening solvent-welded to two 6.1 m 
sections of schedule 40 PVC. This was jetted down inside of the casing 
until it bottomed out at the bottom of the previously augered hole. 
After placement of the well, a mixture of bentonite and water was poured 
into the area between the casing and well up to the top of the casing. 
The purpose of this well was to observe the piezometric head difference 
across the restricting layer. 

4.3 Chemical Applications 
The first application of chemicals to the experimental site 
occurred on 11/12/86. On the day preceding this first application, the 
entire field site was rotary mowed. The grass within the application 
area was cut to a height of approximately 5 cm. All dead grass and 
clippings were raked up and removed from the area prior to application. 
The application area consists of a strip 36.6 m long by 9.14 m wide as 
shown in Figure 4.1. 

Atrazine and alachlor were applied simultaneously on November 12. 
Atrazine was applied in the form of AAtrex (Ciba-Geigy Corp.) which is 
an emulsifiable concentrate containing 0.479 kg AI (active ingredient) /L 
(4 lb/gal) . Alachlor was applied in the form of Lasso (Monsanto 
Agricultural Products Co.) which is also an emulsifiable concentrate 
containing 0.479 kg AI/L. The herbicides were applied using a self- 
propelled boom sprayer with provisions for injecting chemicals directly 
into the water stream (Sumner et al. , 1987) . The boom length on the 
sprayer is 9.14 m. 



50 

The intended application rate of the herbicides was 4.5 kg Al/ha (4 
lb/ac) . One liter of herbicide solution was prepared by mixing 313 mL of 
each of the chemicals with 374 mL of deionized water. The sprayer was 
operated at a velocity of 3.05 m/min moving from south to north along the 
edge of the application area. The chemical solution was injected into 
the water stream of the sprayer using an injection pump calibrated to 
deliver 83 mL/min- The water pressure at the inlet of the sprayer was 
207 kPa. The pressure at the spray nozzles was maintained at 138 kPa. 
Despite the calibration of the injection pump and measurement of sprayer 
ground speed, the chemical solution ran out as the sprayer reached well 
10-09 which is 6.1 m short of the intended end of the application area. 
The application of the herbicides was considered complete at this point, 
and the remaining portion of the application area was left untreated. 

Fourteen 0.35 1 plastic cups were randomly placed within the 
application area. The cups were supported upright in holders at a 
height of approximately 35 cm above the soil surface. The cups were 
used to collect samples of the application solution for analysis of 
chemical concentrations, determination of the depth of water applied to 
the area during application, and uniformity of chemical and water 
application. The use of these cups and holders has been tested and 
found to accurately reflect water application amounts (Stansell et al., 
1982) . 

Immediately after completing the application of the herbicides, the 
sprinkler irrigation system was started and 5.1 cm of water was applied 
over the entire study site. This irrigation was intended to wash the 



51 

herbicides off of the grass foliage and move them into the soil profile 
to minimize volatilization losses from the soil and plant surfaces. 

Bromide in the form of potassium bromide (KBr) solution was applied 
to the application area using the same sprayer on 11/17/86. There was a 
light drizzle of rain during application. Five hundred grams of KBr was 
dissolved in 1050 mL of deionized water and injected into the sprayer as 
described above. This was equivalent to an application rate of 10 kg/ha 
of bromide. The sprayer again was operated at a nozzle pressure of 138 
kPa, and a ground speed of 3.05 m/min. Fourteen collectors in the same 
locations as for the herbicide application were used to collect 
application samples. 

No further chemical applications were made prior to fertilization 
of the grass on 4/16/87. The fertilizer was applied at a rate of 560 
kg/ha of 5-10-15 using a tractor-mounted broadcast spreader. Fifteen 
percent of the nitrogen in the fertilizer was in the form of nitrate 
nitrogen (N0 3 -N) . This is equivalent to applying 18.6 kg/ha of nitrate 
(N0 3 ). 

In an effort to further characterize bromide movement within the 
soil profile and determine flow velocities within the groundwater, an 
additional bromide application and a separate chloride application were 
made. On 4/27/87, KBr was applied to the application area using the 
sprayer utilized during the previous application. The application 
solution was made by dissolving 1.2 kg of KBr in 2.3 1 of deionized 
water. The sprayer made two passes across the application area during 
which it applied approximately 1.6 1 of the KBr solution. This was 
equivalent to an application rate of 17 kg/ha of bromide. Fifteen 



52 

plastic cups were used to catch application solution samples in the same 
manner as previously described. Immediately following application the 
field was irrigated. Irrigations continued daily for a week in order to 
raise the water table and provide plenty of infiltrating water to 
transport the bromide through the soil. 

Prior to the beginning of irrigations on April 27, most of the wells 
on the site were dry. The water table began to rise on May 1 at which 
time 10 1 of solution containing 500 g of potassium chloride (KC1) was 
poured directly into well 07-09 which is near the top of the site. It 
was anticipated that the high concentration of chloride in the solution 
could be followed downslope with the saturated flow and allow for 
determination of water table flow velocities. 

4.4 Sample Collection and Storage 
Samples were collected weekly (Monday) throughout the study. There 
were occasional periods of more frequent sampling (immediately after 
initial applications and during the last week of April through the 
middle of May) . 

The flint glass containers used for well sample collection and 
storage throughout this study did not have teflon lined caps. Therefore, 
a small piece of parafilm (a common laboratory film used for sealing 
beakers and containers to prevent evaporation and contamination) was 
stretched over the mouth of the bottle prior to screwing on the caps. 

Samples of the application solutions of the herbicides and bromide 
were transferred to glass containers and placed in a cooler with ice 



53 

packs immediately after the sprayer had passed completely by the sample 
collector. 

Soil samples were collected in one of two ways: taking 2.5 cm 
diameter by 5 cm long cores vising a soil sampling probe, or using a 5 cm 
diameter stainless steel bucket auger. The soil sampling probe was 
generally used for collection of samples from the top 5 cm of the soil 
profile. When this method was used, 5 to 8 samples would be collected 
from a small area and composited into a plastic sample bag. This would 
represent one final sample for the soil surface at that location. For 
deeper samples the auger was used. The auger was used to remove soil to 
within approximately 5 cm of the desired sample depth. The auger was 
then cleaned to remove traces of soil, and a sample was collected which 
represented soil from a depth of 5 cm above to 5 cm below the target 
depth. The sample was carefully poured from the top of bucket into a 
plastic sample bag and sealed. The number of locations within the 
application area sampled depended on the number of depths to be sampled 
and ranged from 5 to 14. After collection, the soil samples were placed 
into coolers with ice packs for transport to Gainesville. Soil samples 
from outside of the application area were collected several times during 
the study to use as blanks and spikes during soil residue analysis. 

Vacuum was applied to the solution samplers immediately upon 
arrival at the study site (7:00 - 8:00 am) . The vacuum would typically 
be left on for 8 to 10 hours prior to collecting the soil solution 
samples. The vacuum applied to the samplers was approximately 28 cm of 
Hg. Near the end of the day the vacuum would be released and the samples 
would be drawn directly into individual containers. The containers are 



54 

made of polycarbonate with a capacity of 50 mL. The containers had 
polycarbonate screw caps for sealing. Immediately after all soil 
solution samples were collected, the sample containers were placed in a 
cooler for transport. 

The elevation of the water table was measured using a well depth 
indicator with a stainless steel probe. The depth indicator would sound 
an audible alarm when the probe contacted water in the wells. The probe 
was attached to a flat cable with markings in feet and inches, with the 
smallest division being 1/4 in. Water depths were recorded early in the 
morning after arrival. 

After the water table elevations in all wells were measured, the 
wells were pumped out to insure that the water to be sampled was 
representative of the surrounding water in the saturated zone at that 
time. When possible, the well was pumped dry. When the depth of water 
in the well exceeded approximately 0.4 to 0.5 m, the pumping rate was 
less than the rate of flow into the well. When this occurred, the well 
was pumped long enough to insure that 2-3 well volumes had been pulled 
from the well. A small, 12v battery powered, peristaltic pump was used 
to pump the wells and collect samples. 

Samples were collected after all wells had been pumped as described 
above. A rubber stopper was attached to the intake side of the pump 
through aim length of 1.25 cm OD Tygon tubing. The stopper also had a 
piece of Tygon tubing approximately 10 cm in length connected to it for 
attachment to the sampling tube in each well. The stopper was placed in 
a sample bottle and the short piece of tygon tubing was attached to the 
well sample tube. Since the pump was downstream from the sample con- 



55 

tainer, the only components common to each sample collection were the 
rubber stopper and the short piece of Tygon tubing. When a sample was 
collected, the initial water from the well was swirled around the bottle, 
and the bottle was then inverted to remove this rinse water. This served 
to rinse the Tygon tubing, stopper, and bottle. Thus, cross 
contamination between wells was minimized. The sample bottles used for 
the collection and storage of well samples had a capacity of 250 mL. 
Well samples were placed in coolers with ice packs for storage until they 
could be refrigerated. 

All soil samples were frozen as soon as possible after collection 
and maintained at -18°C until preparation for analysis. All water 
samples were refrigerated after collection and maintained at 4°C until 
prepared for analysis. 

4.5 Sample Analysis 
4.5.1 Inorganic tracer analysis 

The concentrations of the three inorganic chemicals of interest 
which were applied during this study were quantified using ion 
chromatography. A DIONEX QIC ion (^hromatograph (IC) was used for all 
analyses. 

A DIONEX AS4A anion separation column was used with an AG4A guard 
column in place to remove organic contaminants prior to reaching the 
separation column. The IC was equipped with a conductivity detector 
connected to an external integrator/recorder. The IC was also equipped 
with an anion micrcmembrane suppressor to reduce the background 



56 

conductivity of the eluant and thus allow for greater sensitivity to the 
conductivity changes caused by the presence of anions in the sample. 

There was no automatic sampling device with this instrument, so 
each sample was injected by hand. Approximately 1 mL of sample was 
loaded onto a sample loop prior to injection onto the column. The 
sample loop retained approximately 0.25 mL of the sample with the rest 
being discarded to a waste line. Between injections, the sample loop 
was flushed with 1 mL of deionized (DI) water. Prior to the guard 
column, the sample was passed through two filters rated at 30 and 5 
microns, respectively, to remove solids from the sample. These filters 
were an integral part of the IC and the filter elements were changed 
whenever the system pressure was observed to be increasing above nominal 
levels. 

The eluant used in these analyses was 2.2 millimolar (mM) sodium 
carbonate (0.933 g Na 2 a>3 in 4 1 DI H 2 0) and 0.75 mM sodium bicarbonate 
(0.25 g NaH00 3 in the same 4 1 of DI water) . This solution was made 
fresh as needed. All references herein to deionized (DI) water refer to 
water that had been passed through carbon filters and ion exchange resins 
to produce DI water with a resistivity of at least 1 megohm. The 
regenerant solution was 0.025 N sulfuric acid (2.8 mL of concentrated 
H 2 S0 4 in 4 1 of DI water) . The regenerant solution was passed through 
the micromembrane suppressor to reduce background conductivity. 

In normal operation the eluant flowrate was approximately 2.0-2.5 
mL/roin, and regenerant flowrate was approximately 4 ml/min. The 
pressure at the discharge of the pump at the stated eluant flowrate was 
approximately 8,274 kPa (1200 psi) . 



57 

A single injection could reveal the presence and concentrations (if 
sta n dards were prepared) of the following anions (in order of increasing 
retention times) : fluoride (F~) , chloride (Cl~) , bromide (Br~) , nitrate 
(N0 3 ~) , phosphate (P0 4 3 ~) , and sulfate (S0 4 2 ") . 

In ion chromatography, the retention times of the various anions are 
dependent upon total ionic strength of the sample, ionic strength of the 
individual ions, eluant flowrate, and temperature of the sample and 
eluant. Ihus, the retention times could, and did, vary. When the 
integrator was set to calculate the areas under the peaks, it was found 
that the automatic selection of the beginning and ending points of peaks 
was not consistent. A small change in retention time could 
significantly alter the reported area for a given ion even when using 
duplicate injections. Better results were achieved by setting the 
integrator to report maximum peak height. This value was not sensitive 
to variations in the starting and stopping times of a peak as long the 
baseline was relatively stable. 

Initial injections of water table samples from the experimental 
site revealed the presence CI" and S0 4 2 ~ in all samples, and relatively 
high concentrations of N0 3 " in samples from well IDW. Standards were 
prepared using DI H 2 and oven-dried quantities of certified or primary 
standard grades of potassium chloride, potassium bromide, potassium 
nitrate, and potassium sulfate. The standards initially included CI" 
and S0 4 2 " because of their presence in the samples. They were not used 
to quantify the concentrations of either species during initial sample 
analysis. When Cl~ was introduced into the water table as a tracer late 
in the experiment, sample concentrations of Cl~ were quantified. It was 



58 

observed that the response of the chromatogram to Cl~ was approximately 
twice the response to the other anions of interest. Thus, standards were 
prepared with Cl" at half of the concentrations of the other species so 
that the responses of all species would be similar and on scale 
simultaneously. Standards were prepared with concentrations of the main 
species of interest ranging from 0.01 to 100.0 mg/L. A typical standard 
would contain 0.5 mg/L Cl" and 1.0 mg/L of Br", N0 3 ~, and S0 4 2 ". It 
should be noted here that all references to nitrate (N0 3 ~) indicate 
concentrations of N0 3 " and not N0 3 -N (nitrate-nitrogen) . 

Samples were taken from the refrigerator and allowed to come to 
room temperature prior to injection. No sample preparation was 
performed. The analysis protocol typically followed the pattern of 3 
standards, 10 samples, 3 standards, etc. The standards used were 
selected based upon the concentrations of the analytes being observed in 
the samples. Thus, standards were selected to bracket the observed 
values as closely as possible. Occasionally a sample would be reinjected 
to observe the repeatability of the analysis procedures. Samples from 
wells located away from the experimental site, including the deep well 
which supplies irrigation water to the farm, were regularly analyzed as 
field blanks (no Br" would be expected; however, Cl", N0 3 ~, and S0 4 2 " 
could be present) . 

A piecewise linear fit to each sequence of injected standards was 
used to generate a function of the form: concentration = f (peak height) . 
The piecewise fit was selected instead of linear regression because for 
some species (particularly high concentrations of Cl") a linear 
regression resulted in a negative y-axis intercept. This implies that 



59 

the species can generate a noticeable positive response and yet be 
calculated to have a negative concentration. It is expected that a 
positive intercept would be a normal result, i.e. there is some threshold 
concentration of the species required before the detector indicates a 
response. The piecewise fit insured that any positive response would 
result in a positive calculated concentration. Conversations with an 
organic chemist at the Pesticide Residue Laboratory of the USDA-ARS 
Southeast Watershed Research Laboratory indicated that the piecewise fit 
would be an acceptable method of generating a standard curve (Marti, 
1987) . 

The standard curves generated on either end of the analysis of 10 
samples were used to compute the concentrations of the species in the 
samples. The resulting concentrations were then linearly weighted based 
on sample position with respect to either set of standards, i.e. the 
first sample following the standards would be weighted towards the curve 
generated by the preceding standards and only slightly effected by the 
standard curve generated 9 samples later. The sample midway between 
standards would be evenly weighted between the two standard curves. In 
general, there was little drift in the instrument and the standard 
curves did not vary much over a period of hours. 

4.5.2 Herbicide residue extraction 

Analysis of samples for residues of atrazine and alachlor required 
different extraction procedures for soil and water samples. Once the 
extractions were completed, all samples were analyzed on a common gas 
chrcmatcgraph with identical operating conditions. 



60 

There are many published methods for preparing samples for 
determination of atrazine and alachlor residues (e.g., Rohde et al. , 
1981; Voznakova and Tatar, 1983) . In general, the procedures are rather 
detailed and costly in terms of time and supplies. Dr. Willis Wheeler, 
Director of the Pesticide Residue Laboratory in the Food Science and 
Human Nutrition Department at the University of Florida, Gainesville, 
indicated that a simpler method of sample preparation might be developed 
which would be satisfactory in terms of the recovery of the pesticide of 
interest while being more economical in both the time required for sample 
preparation and expense (Wheeler, 1987) . The more steps involved in the 
extraction and preparation of a sample, the more likely it is that 
something can go wrong and the residues lost. With these factors in 
mind, methods for extraction of the two herbicides from soil and water 
samples were developed which were in keeping with the budgetary 
constraints of the project and the time constraints on the analyst. 

Extraction of the herbicide residues from water samples was done 
using a liquid-liquid extraction procedure. One hundred fifty 
milliliters of the sample water were placed in a quart mason jar along 
with 150 mL of pesticide-grade hexane. The jar was sealed with a 
standard mason jar lid which had been lined with a self-adhesive teflon 
film. The jar was agitated on a shaker table for thirty minutes after 
which the contents of the jar were poured into a labeled 500 mL 
separatory funnel. The sample water was drained through the separatory 
funnel back into the jar, leaving the hexane in the funnel. Another 150 
mL of hexane was added to the jar and it was again agitated for 30 
minutes. The contents of the jar were again added to the funnel. The 



61 

water was then drained and discarded. The hexane was filtered into a 500 
inL boiling flask through sodium sulfate to remove any remaining water. 
Next, the boiling flask was placed on a rotary evaporator and the 
reduced in volume to near dryness. A few mL of fresh hexane was then 
added to the flask and swirled to wash the sides of the flask. This 
hexane was placed in a 10 mL volumetric flask. A few mL of hexane was 
again added to the boiling flask and the swirling repeated. After two 
rinses of the flask, the hexane in the volumetric flask was brought to a 
final volume of 10 mL. The finished sample was stored under 
refrigeration in test tubes with teflon lined screw closures. The 
initial sample volume of 150 mL and the final extracted volume of 10 mL 
indicate that this procedure concentrated the residues by a factor of 15. 
Blanks taken from remote wells as well as DI water were extracted as a 
check on the procedure. Spiked samples of well water and DI water were 
extracted to determine the recovery levels of the herbicides using the 
procedure described above. 

Soil samples were removed from the freezer and allowed to air dry 
for one day prior to extraction. This was done to allow excess moisture 
to evaporate since the acetone used in the extraction process can absorb 
water which will not be removed by filtration through sodium sulfate as 
was done for the hexane extractions. Acetone was chosen as the solvent 
for soil extractions because it was observed that the water in the soil 
samples, or some other factor, tended to repel hexane. It did not 
appear that the hexane was adequately dispersing the soil particles. 

A fifty gram portion of the air-dried soil sample was weighed and 
placed into the mason jar. One hundred mL of pesticide grade acetone 



62 

was then added to the soil sample. The sample was agitated for 30 
minutes and the acetone was carefully decanted into a separate labeled 
glass container. Another 100 mL portion of acetone was added to the soil 
and the agitation was repeated for another 30 minutes. The acetone was 
poured into the container with the previous portion. Several rinses of 
acetone were poured through the extracted soil to remove any acetone 
left from the extractions. The acetone rinses were added to the acetone 
from the extractions and the entire volume was vacuum filtered through a 
47 mm diameter filter with a 0.45 micron pore size. The filtrate was 
then reduced to near dryness on the rotary evaporator and brought to a 
final volume of 10 mL in acetone. Soil samples taken from outside the 
study area were extracted as field blanks and were spiked with known 
quantities of the herbicides and extracted as a check on the extraction 
and analysis procedures. 

A subsample of approximately 10 g of each soil sample was taken to 
determine the soil-water content so that the concentrations of the 
herbicides in the soil could be reported on dry weight basis. Soil 
samples for soil-water content determinations were dried for 24 hours in 
an oven at 100°C. 

4.5.3. Herbicide residue analysis 

All herbicide residue samples were analyzed on a Hewlett Packard 
model 5840A gas chromatograph (GC) . The GC was equipped with a 35 
position autosampler and a model 5840 GC terminal (integrator) . The GC 
had both electron capture (EC) and nitrogen-phosphorus (NP) detectors. 
EUring the early phases of the study (methods development) , the electron 



63 

capture detector was used. This detector had a high sensitivity to 
alachlor and a lower sensitivity to atrazine. However, when field 
samples were analyzed, as opposed to DI water spiked with pesticides, it 
was observed that the EC detector was sensitive to many naturally- 
occurring compounds (the EC detector is sensitive to chlorine molecules) 
which had been extracted along with the pesticides. Thus, the output was 
so crowded with peaks that the quantities of atrazine and alachlor 
present could no longer be reliably determined. The NP detector is very 
sensitive to molecules of nitrogen and phosphorus and insensitive to 
almost anything else. When the samples were analyzed on this detector, 
the chromatograms were much cleaner and atrazine and alachlor were often 
the only visible peaks. The NP detector is somewhere between 5 and 10 
times more sensitive to atrazine than it is to alachlor due to the 
greater number of nitrogen atoms in the structure of atrazine. The NP 
detector was operated with detector gas flowrates of air at 50 ml/roin and 
hydrogen at 3 ml/min' Ihe carrier gas was helium with a flowrate of 30 
ml/min. 

The column used in the GC was 1.83 m long with a 4 mm ID. The 
column was packed with 3% OV-17 (50% phenyl, methyl silicone) on 100/120 
mesh Gas Chrom Q. The operating conditions were: injector port 
temperature set at 240°C, column temperature at 200°C, and detector 
temperature at 300°C. 

The sample protocol used with the auto sampler was usually of the 
form: 4 standards, 1 solvent blank, 10 samples, 4 standards, 1 solvent 
blank, 10 samples, 4 standards, 1 solvent blank. The solvent blank was 
used to help prevent carryover from the last (and highest) standard into 



the following sample. Carryover was normally not a problem, but the 
solvent blank would often show some trace of the herbicides. 

Working standards of the two herbicides were prepared using 
reference standards provided by the EPA in Research Triangle Park, North 
Carolina. The initial standard was made by dissolving 10 mg of the 
reference into 10 mL of solvent yielding a concentration of 1 mg/mL 
(1000 mg/L) . This standard was then used as the basis for subsequent 
working standards. The standards were prepared with the concentration 
of alachlor being 5 times greater than the concentration of atrazine so 
that the detector response to the chemicals would be nearly equal. 
Working standards of atrazine ranged from 0.01-500.0 mg/L with the 
concentrations of alachlor generally being 5 times higher as noted above 
(very high atrazine standards did not include alachlor) . The 0.01 mg/L 
standard of atrazine represented the smallest concentration which was 
detectable without pushing the limits of machine and detector. The 
autosampler injected a volume of 4.8 juL- So an injection of 4.8 /iL of 
0.01 mg/L atrazine represented a lowest detectable mass of 10 nanograms 
of atrazine. 

Whenever new standards were prepared, they were compared with the 
previous standards to assure similar response and that no mistakes had 
been made during the dilutions. Two sets of standards were made. One 
set was made in acetone for use in analyzing the soil sample extracts. 
The other set was made in hexane for use in analyzing the water sample 
extracts. Comparisons between the two sets of standards revealed no 
differences in response, but recommended procedures (Wheeler, 1987) 
dictate that the standards be made up in the same solvent as the samples. 



65 

The integrator available with the GC did not work properly. In 
order to quantify the amount of herbicides present in the samples, all 
peak heights were hand measured (a peak height mode, as used with the IC 
was not available on this integrator) . 

A linear regression was performed from the standards to give a 
relationship of the form: concentration = f (peak height) . There were no 
problems with negative intercepts in these regressions as discussed in 
the section on ion chromatography. Sample concentrations were computed 
with the regression equations calculated from standards on either side of 
the samples. In order to account for drift in the detector response, the 
final concentrations were calculated by linearly weighting the 
concentrations computed from each set of standards as discussed 
previously. The NP detector had a tendency to drift and this correction 
for drift was usually needed. 



CHAPTER 5 
MODELING THE EXPERIMENTAL SITE 

Two models were selected for the purpose of simulating the movement 
of the chemicals applied to the experimental site. The models selected 
for use were the Pesticide Root Zone Model, PRZM (Carsel et al., 1984) , 
version 2, and Groundwater Loading Effects of Agricultural Management 
Systems, GLEAMS (Leonard et al., 1987), version 1.8.53. These models 
were selected because they were designed specifically to simulate 
pesticide transport through the root zone of a crop and reflect 
influences of agricultural management practices such as tillage. 
Reference to the manual for GLEAMS includes the documentation for the 
CREAMS model (Knisel, 1980) from which GLEAMS was derived as well as the 
supplementary GLEAMS user manual which is provided with the model code 
and describes the differences in input data sets between CREAMS and 
GLEAMS. 

This chapter will discuss the selection of input values used in the 
models. Only the parameters relevant to the results will be discussed. 
There are parameters in both models for which values were selected, but 
which had no bearing on the results to be presented. For example the 
models utilize the Universal Soil Loss Equation, USLE, (Wischmeier and 
Smith, 1978) for prediction of soil erosion, however, in the simulations 
surface runoff was never predicted and therefore the parameters chosen 
for use with the USLE did not affect the reported results. There are 

66 



67 

many parameters such as rainfall, pesticide properties, and soil 
characteristics which are common to both models. Thus, the selection of 
common parameter values will be discussed first, followed by selection of 
values for parameters which are unique to each model. 

5.1 Selection of Common Input Parameter Values 
5.1.1 Rainfall and irrigation 

Both models use daily rainfall data. The formats of the rainfall 
files for the models are different and PRZM reads in values in 
centimeters while GLEAMS reads in values in hundredths of an inch. 

The data used to create the precipitation files came from a 
combination of two rain gages. As stated previously, one weighing rain 
gage is centrally located within the experimental site. This rain gage 
was installed in June of 1986. A second rain gage is located 
approximately 300 m north-northeast of the experimental site. This gage 
is a tipping bucket type rain gage that is connected to a data logger. 
All rainfall data used in the simulations came from this off -site gage 
since the data from it was already on a computer and available. The 
rainfall amounts recorded in both gages agreed well. The gage located on 
the site was used to supplement the rainfall record with amounts of 
irrigation applied to the site. This rainfall plus irrigation file was 
then manipulated into the units and formats required by each model, and 
checked to insure that the two formats contained identical information. 



68 

5.1.2 Soil properties 

Both models vise the Soil Conservation Service (SCS) curve number 
method (USDA-SCS, 1972) to partition rainfall between runoff and 
infiltration. The runoff curve number was selected from Table 9 in the 
FRZM manual and was chosen based on the assumption that the bahia grass 
cover could be assumed to be similar to a pasture in good condition 
without contouring (Lakeland sand is classified as being in hydrologic 
soil group A) . 

Physical descriptions of the soil profile were obtained from two 
sources. Hook (1985) measured properties of the soil in another section 
of the farm on which the experimental site is located. The data 
collected by Hook (1985) were measured within the top 120 cm of the soil 
profile. The soils on the farm are uniformly classified as either a 
Lakeland sand or Bonifay sand. The primary differentiation between the 
soils is the depth to the restricting layer with the Lakeland soil being 
deeper than the Bonifay. Soil physical properties such as water 
retention and hydraulic conductivities are considered to be the same 
between these soils on this farm. The second source of data used to 
determine soil properties is a soil characterization report for Florida 
soils prepared by the Soil Science Department at the University of 
Florida (Carlisle et al., 1978). This report contains characterization 
data for a Lakeland sand. 

The properties that were selected to represent the soil profile of 
the experimental site are summarized in Table 5.1. The average depth to 
the restricting layer for the three wells within the application area is 
2.62 m and this was chosen to represent the profile depth. 



69 



Table 5.1 Soil properties used in simulations. 



Depth 


Carbon 


Bulk 
Density 


Conductivity 

\ iiy ill y 


Water Content m 
Effective Field 
Saturation Capacity 2 


Wilting 

r\J±l 1 L, 


0- 13 


0.55 


1.45 


13.3 


33.0 


10.0 


2.1 


13- 20 


0.28 


1.60 


13.3 


33.0 


10.0 


2.1 


20- 51 


0.08 


1.58 


22.9 


33.3 


10.0 


2.1 


51-102 


0.04 


1.59 


36.5 


33.8 


9.0 


2.1 


102-262 


0.03 


1.59 


45.0 


34.0 


9.0 


2.1 



1 Data from Hook (1985) and Carlisle, et al. , (1978). 
2 Field gravimetric water content following 48 hours of drainage under 
plastic following 16 hours of flooding. 
3 Water content at 15 bars tension. 

5.1.3 Pesticide chemical properties and applications 

The pesticide chemical properties required by the models are 
solubility, partition coefficient, degradation rates on foliage and in 
the soil, and a coefficient of plant uptake. The models may require 
these parameters in slightly different forms, but the basic information 
required is the same. In addition to the model manuals, the manual for 
LEACH (Dean et al., 1984) was a good source of estimates of the required 
parameters. No exhaustive search of possible values of parameters for 
each chemical was undertaken. At this point in the study, the models 
were run using values that a user could obtain from the manuals. This 
may result in less than optimum performance of the models, but probably 
reflects the way in which a typical user would apply them. 

Reported properties of pesticides are often widely variable. For 
properties such as partition coefficient and degradation rates, the 



70 

variability is understandable and expected. However, even water 
solubility was found to vary by a factor of two. The solubility of 
atrazine is not available from the PRZM manual. The CREAMS manual lists 
a value of 33 mg/L. However, the Agrichemical Handbook (1983) gives a 
value of 70 mg/L at 20°C. The solubility of alachlor is given in the 
PRZM manual as 220 mg/L at 20-25°C. The CREAMS manual reports a value of 
242 mg/L at an unspecified temperature. The Farm Chemical Handbook 
(1984) does not report a value for alachlor and the Agrichemical Handbook 
(1983) gives the solubility of alachlor in water as 148 mg/L at room 
temperature. The final values selected for all chemical properties are 
shown in Table 5.2. 

The solubilities of bromide, nitrate, and chloride listed in Table 
5.2 are considerably less than the actual solubilities. However, 
solubilities of ionic species are seldom listed for reference. The 
values in Table 5.2, while less than the actual solubility values, are 
sufficiently high such that solubility was not a limiting factor in the 
mobility of the chemicals. 



Table 5.2 Chemical properties used in simulations. 

Property Atrazine Alachlor Bromide Nitrate Chloride 

Solubility 33 148 800 800 800 

(mg/L) 

Partition coef . , 163 268 

Kqc (cmVg) 

Half-life 78 18 999 999 999 

(days) 

Plant Uptake Coef. 1.0 1.0 1.0 1.0 1.0 

0.65 0.52 

0.0 0.0 0.0 0.0 0.0 



71 

The laser manuals for the models contain tables drawn from many 
sources which provide values of the partition coefficient for many 
pesticides. The LEACH manual also contains a number of tables. The 
tables in CREAMS and LEACH are more extensive than those in FRZM. 
Different tables within the same manual may give differing values. Table 
values often list coefficients of variation (CV) on the order of 50-130 
percent. FRZM presents equations by which the organic carbon partition 
coefficient, Kqq, can be calculated if solubility or the octanol^water 
partition coefficient, K^, is known. 

The partition coefficients of alachlor and atrazine were not 
measured on this soil and thus values were chosen from the sources 
described above. A Kqq value for atrazine of 163 cm 3 /*? with a CV of 49% 
was found in the LEACH manual. No listing of a Kqq was found for 
alachlor. FRZM did list a value for logCK^) which was 2.78. FRZM 
presents a relationship between log(K ow ) and Kq,-. which is 

log Kqc = 1.00 (log - 0.21 5.1 

Using this equation a Kqq of 371 was calculated for alachlor. LEACH 
lists a value for alachlor as 434, using this and equation 5.1, a 
value of 268 was calculated. Since alachlor has been reported as one of 
the pesticides commonly found in groundwater, it was decided to choose 
the lower Kq,-. value for use in the model as this would tend to cause 
higher predicted leaching losses. 

The inorganic tracers bromide, chloride, and nitrate were assumed to 
act as non-adsorbed pesticides. Thus they were assigned a Kqq value of 
zero. 



72 

The degradation rate constant of atrazine in soil was given in the 
LEACH manual as ranging from 0.0149 to 0.0063 days" 1 , corresponding to a 
half-life of between 46 and 110 days respectively. The midpoint value 
of 78 days was chosen for use in these simulations. A degradation rate 
constant of 0.0384 days -1 , which corresponds to a half-life of 18 days, 
was given for alachlor in both the CREAMS and LEACH manuals. 

The coefficient of plant uptake of pesticides with transpiration was 
determined using the relationship given in the PRZM manual in which the 
uptake factor is a function of Kq^ and is given as 

UPTKF = 0.784 exp - [(log - 1.78) 2 /2.44] 5.2 
where UPTKF = plant uptake efficiency factor. 
Using the for alachlor of 434 as discussed above, the uptake factor 
was calculated to be 0.52. Back calculating a for atrazine, based 
upon the chosen K^-. of 163 using equation 5.1, yields a of 264. 
Using this value in equation 5.2 gives a plant uptake factor of 0.65. 

The models were also run using uptake factors for both atrazine and 
alachlor of 0.0 and 1.0 in order to assess the sensitivity to these 
factors. GLEAMS currently recommends that in the absence of well defined 
uptake coefficients, a value of 1.0 should be used. Uptake factors for 
the inorganic tracers were assigned values of both 0.0 and 1.0 to observe 
the sensitivity of these non-adsorbed chemicals to plant uptake. 

The dates and amounts of the chemicals applied to the soil surface 
of the experimental site were entered into the models in a direct manner, 
ie. there were no calculations or transformations required to convert the 
actual values into model parameters. The application rates used in the 
simulations correspond to the intended application rates as opposed to 



73 

the measured rates. The dates and application rates used as model inputs 
are shown in Table 5.3. The applications were assumed to reach the soil 
surface with no chemical residues remaining on plant surfaces. The 
amount applied was further assumed to be uniformly incorporated into the 
top 1 cm of the soil. 

5.1.4 Length of simulations 

Both models simulated the two year time period from 1/1/86 through 
12/31/87. This was done to allow the models to overcome any effects of 
initial conditions prior to beginning simulation of chemical 
applications. Continuing simulations for six months beyond the end of 
field data collection would show if any significant leaching would likely 
have occurred after sampling was terminated. 



Table 5.3 Chemical applications summary. 



Date 


Chemical Applied 


Application Rate 
(kg/ha) 


11/12/86 


Atrazine 


4.9 




Alachlor 


4.9 


11/17/86 


Bromide 


10.0 


4/16/87 


Nitrate 


18.6 


4/27/87 


Bromide 


17.0 



5.2 Parameters Unique to FRZM 
5.2.1 Evapotranspiration prediction. 

PRZM uses either measured daily pan evaporation or daily mean 
temperature to determine evaporation from the soil and plant surfaces, 
and transpiration by the crop. In these simulations, daily pan 



74 

evaporation values recorded at a weather station located on the 
University of Georgia Coastal Plain Experiment Station campus were used. 
This weather station is located approximately 8 km from the research 
site. PRZM requires a pan coefficient which is a multiplier used to 
adjust pan readings to represent total daily potential evapotranspiration 
(PET) . A pan factor of 0.75 was selected from a figure provided in the 
manual showing pan factors as related to geographic location. 

Two additional parameters required for evapotranspiration (ET) 
prediction are the depth in the soil profile from which evaporation can 
occur, and the depth of rainfall and irrigation that can be stored on the 
plant canopy (interception storage) . A figure in the manual indicates 
that for southern Georgia, the depth of soil contributing to evaporation 
is approximately 25 cm. Data pertaining to the interception storage 
capacity of bahia grass was not included in the manual. Values presented 
in the manual for various crops ranged from 0.0-0.3 cm. Assuming that 
the storage capacity of bahia grass is relatively small, a value of 0.05 
cm was assumed. 

5.2.2 Crop related parameters 

Characteristics of the crop(s) grown during the simulation period 
are defined by parameters related to rooting depth, maximum canopy 
coverage, and dates for emergence, maturation, and harvest. A rooting 
depth for bahia grass was difficult to obtain. Most references to bahia 
grass simply stated that it is a deep or very deep rooted plant. Based 
on such generalizations and without more specific information, an active 
rooting depth (depth containing 90% of root mass) of 91 cm (3 ft) was 



75 

selected. Based on observations at the research site, a maximum areal 
coverage of the bahia grass foliage was estimated as 98 percent. 

Dates of crop emergence, maturation, and harvest are used in the 
model to define the development and growth of roots. Roots are assumed 
to begin growing at crop emergence and reach the maximum rooting depth at 
plant maturation. The roots stay at maximum depth until harvest when 
they are set back to zero. These parameters are somewhat difficult to 
define in relation to a perennial crop such as bahia grass. In view of 
the use made of the selected dates, some artificial dates for emergence 
and harvest could be selected. Since the emergence date really defines 
the beginning of plant transpiration as a component of the water balance, 
an emergence date of March 1 was selected to correspond to the time when 
the grass begins to "green up" in the spring. Since the majority of the 
roots already exist when the grass begins active growth in the spring, a 
maturation date of April 1 was selected to reflect water extraction from 
the full maximum rooting depth shortly after growth for the new year 
begins. The harvest date was chosen to be December 1 which reflects the 
approximate date when the grass turns brown in the winter and active 
growth ceases. 

5.2.3 Soil related parameters 

PRZM will allow simulation of a larger section of the profile than 
just the root zone. A soil core depth of 2.62 m was chosen to represent 
the average profile depth beneath the application area. The model will 
allow the user to select the number of aanpartments used to represent the 
profile. Too many compartments will increase simulation times and too 



76 

few will lead to increased numerical errors in the solution of the 
equations used within the model. The manual suggests that no fewer than 
30 compartments be used. Thirty-five compartments were selected because 
this gives a cxjttpartment depth that is one-half the cxxrpartment depth in 
GLEAMS for a root zone depth of 91 cm (see section 5.3.3) . 

FRZM and GLEAMS both normally operate on the assumption that any 
soil water in excess of field capacity will drain into the next lower 
compartment. This drainage will continue until any excess water drains 
from the bottom of the simulated profile. This drainage is assumed to 
occur within one day. In the PRZM model, this type of drainage is called 
'free drainage' . FRZM has a feature to allow selection of a drainage 
rate parameter that serves to slow down the percolation process. This is 
an empirical constant and requires calibration. A figure is provided in 
the manual whereby the user can make a first cut selection of the 
drainage parameter based upon the soil type (sand, clay loam, etc.) and 
the number of simulation compartments selected as described above. 
Observations at the experimental site clearly show that there is a delay 
of 2-4 days from the time excess water is applied to the time that the 
water table responds. Therefore a few simulations were made using the 
restricted drainage option to compare with the experimental data. The 
drainage rate parameter for this case was selected from the graph in the 
manual for a sand soil with 35 compartments to be 2.63 day -1 . 



77 

5.3 Parameters Unique tn nrrAMS 

5.3.1 Evapotranspiration prediction 

Potential evapotranspiration (PET) is computed in GLEAMS based on 
mean monthly maximum and minimum temperatures and mean monthly solar 
radiation. The mean monthly temperatures and solar radiation data were 
obtained from the same weather station that was used for the pan 
evaporation data input to PRZM (section 5.2.1). GLEAMS also 
differentiates between soil evaporation and plant transpiration. A soil 
evaporation parameter is required as a model input and the manual gives a 
suggested value of 3.3 (no units) for this parameter for sands. 

5.3.2 Crop related parameters 

Crop related parameters which are different from PRZM are leaf area 
index (LAI) , and winter cover factor. The LAI data are used to reflect 
changes in plant transpiration with stage of growth. In keeping with the 
observed behavior of bahia grass discussed in section 5.2.2, the LAI of 
bahia grass was assumed to be zero until March 1 of each year and to 
return to zero on December 1st. The maximum LAI for bahia grass was 
assumed to be similar to the LAI for pasture as presented in the CREAMS 
manual which had a maximum LAI of 3.0. The LAI for the bahia grass was 
assumed to reach 2.0 on April 1 and to reach 3.0 May 1 where it remained 
until it began to decline again to a value of 2.0 on Nov. 1. 

The winter cover factor is used to moderate evaporation from the 
soil surface during the winter due to presence of a bare surface or cover 
crops. Since the soil surface at the site was not bare during winter, 
the recommended value of the winter cover factor of 0.5 was used. 



78 



5.3.3 Soil related parameters 

GLEAMS was designed to simulate only the root zone of a crop. The 
vadose zone below the crop is ignored. Thus, the soil profile 
description input to GLEAMS covers only the 91 cm root zone as opposed to 
the 2.62 m profile input to FRZM. There are no soil parameters which are 
significant to the results presented in this work which differ from those 
discussed in section 5.1.2. In order to highlight one of the major 
differences between the two models, a short discussion of the 
computational layering of the soil profile in GLEAMS will be presented 
here. GLEAMS divides the crop root zone into seven computational layers. 
The first layer is defined to be 1 cm in depth and the other 6 layers are 
equal in thickness (layer 2 is 1 cm less than layers 3-7) and together 
account for the rest of the root zone. The 1 cm upper layer is 
considered an important feature of CREAMS and GLEAMS because it increases 
the sensitivity of erosion and runoff to chemical concentrations near the 
soil surface. FRZM divides the entire profile into a specified number of 
compartments of equal depth. For the case presented here, GLEAMS layers 
would be 15 cm thick after the surface layer (91 cm / 6 layers) . For 
PRZM with a 2.62 m profile and 35 compartments, each compartment will be 
7.5 cm thick. Thus, FRZM may be better able to describe effects of 
actual soil layering since properties of adjoining layers would not be 
averaged over as large a layer as in GLEAMS. However the surface layer 
in FRZM will also be 7.5 cm thick and pesticide storages will be averaged 
over this whole layer effectively lowering the mass at the surface which 
is available for transport with eroded sediment and runoff water. 



CHAPTER 6 
RESULTS AND DISCUSSION 



There are many areas in this study for which results can be 
reported. The approach used in this chapter is to report significant 
results relative to each area such as sample collection, sample analysis, 
simulation results, etc. , and then relate the individual results where 
appropriate. A summary of activities at the field site is presented in 
Table 6.1 for reference. 

6.1 Data Collection 

The site was visited on at least a weekly basis throughout the 
study. At each visit, samples were collected from the monitoring wells 
and soil solution samplers. The water table elevation was also recorded 
for every monitoring well on the site and for several wells located 
adjacent to the study area. Soil samples were collected on an irregular 
basis during the study. 

The procedure used for collecting samples from the groundwater 
worked well in general. Sand intrusion into a few of the wells, however, 
was a consistent problem. It was possible in some instances to fill the 
250 mL sampling bottle with sand. This problem was overcome by slowly 
raising the bottom of the sampling tube in order to stay above the sand. 
These wells were pumped out using a 1.6 cm diameter suction line to 
remove acajmulated sand as needed. The problem could have been avoided 

79 



80 



Table 6.1 Chronological summary of field site activities. 



Date 



11/12/86 



11/17/86 



11/18/86 



Event 



Apply 4.9 kg/ha of atrazine and alachlor, apply 5.1 cm 

of irrigation after herbicide application. 

Apply 500 g of KBr (10 kg/ha Br") , collect soil solution 

samples. 

Collect soil solution samples. First soil samples taken 
from top 5 cm in application area. 



11/24/86 


Collect soil 


solution and soil samples. 


12/01/86 


Collect soil 


solution samples. 


12/08/86 


Collect soil 


solution samples, apply 5.1 cm of 




irrigation. 




12/15/86 


Collect 


soil 


solution and well samples. 


12/22/86 


Collect 


soil 


solution, well, and soil samples. 


12/29/86 


Collect 


soil 


solution and well samples. 


1/05/87 


Collect 


soil 


solution and well samples. 


1/12/87 


Collect 


soil 


solution and well samples. 


1/19/87 


Collect 


soil 


solution and well samples. 


1/26/87 


Collect 


soil 


solution and well samples. 


2/02/87 


Collect 


soil 


solution and well samples. 


2/09/87 


Collect 


soil 


solution, well, and soil samples. 


2/16/87 


Collect 


soil 


solution and well samples. 


2/23/87 


Collect 


soil 


solution and well samples. 


3/02/87 


Collect 


soil 


solution and well samples. 


3/09/87 


Collect 


soil 


solution and well samples. 


3/16/87 


Collect 


soil 


solution, well, and soil samples. 



81 



Table 6.1 Continued. 
Date Event 
3/23/87 Collect soil solution and well samples. 

3/31/87 Collect soil solution and well samples. 

4/06/87 Collect soil solution and well samples. 

4/13/87 Collect soil solution and well samples. 

4/16/87 Apply 560 kg/ha of 5-10-15 to entire study site. 

4/20/87 Collect soil solution and well samples, Very few wells 

have water. 

4/27/87 Collect soil solution and well samples, Apply 2.5 cm of 

irrigation, apply approximately 17 kg/ha Br", apply 2.5 
cm of irrigation. 

4/28/87 Collect soil solution and well samples, apply 4.2 cm of 

irrigation. 

4/29/87 Collect soil solution and well samples, apply 3 cm of 

irrigation. 

4/30/87 Collect soil solution and well samples, apply 3 cm of 

irrigation. 

5/01/87 Collect soil solution and well samples, pour 10 1 of 

solution containing 500 g KCl into well 07-09, apply 2.7 
cm of irrigation. 

5/02/87 Collect soil solution and well samples, apply 3.2 cm of 

irrigation. 

5/03/87 Collect soil solution and well samples, apply 1.9 cm of 

irrigation. 



82 



Table 6.1 Continued. 



Date 



Event 



5/05/87 

5/08/87 

5/11/87 

5/13/87 
5/18/87 
5/25/87 
6/01/87 



Collect soil solution and well samples, apply 0.9 cm of 
irrigation. 

Collect soil solution and well sanples, apply 1.9 cm of 
irrigation. 

Collect soil solution and well sanples, apply 5.2 cm of 
irrigation. 

Collect soil solution and well sanples. 
Collect soil solution and well samples. 
Collect soil solution, well, and soil sanples. 
Collect soil solution and well samples. 



by using commercially available PVC well screening in the monitoring 
wells. 

No problems were observed with the operation of the soil solution 
samplers. The volumes of water collected during sampling varied from 
to greater than 50 mL. Even though the samples were pulled back to the 
instrument trailer through as much as 35 m of 0.64 cm ID polypropelene 
tubing, the vacuum applied to the sample lines was sufficient to pull 
very small droplets of water along the wall of the tube and into the 
containers. No carryover of the tracers between sampling periods was 
observed. 

The tensiometers were difficult to maintain properly on the weekly 
sampling schedule. Even when the tensiometers were working, their 
usefulness was limited by the short length of time during which they 



83 

could be observed. If an irrigation was initiated early in the morning, 
a few of the 30 cm tensiometers might have responded by the end of the 
day. Weekly observations of soil-water tension were insufficient for the 
purpose of calculating the flux of water through the unsaturated zone. 

6.2 Sample Analysis 

Water samples collected from the monitoring wells were analyzed for 
residues of atrazine and alachlor and concentrations of the inorganic 
tracers utilized in this research (Br", N0 3 ", and Cl~) . The volume of 
samples from the solution samplers ranged from to more than 50 mL, with 
the average sample volume being approximately 10 mL. Water samples from 
the soil solution samplers were generally analyzed for the inorganic 
tracers only. Most of the solution samples were not extracted for 
residues of the herbicides due to the small sample volumes. 
Approximately 50 of the soil solution samples were extracted and analyzed 
for residues of the herbicides. Soil samples were extracted and analyzed 
for residues of atrazine and alachlor. The presence of alachlor and 
atrazine residues in the soil samples was confirmed in a composite of 
soil sample extracts that was analyzed using gas c±romatography/mass 
spectrometry (GC/MS) . 

During extraction of water samples with hexane, an emulsion of 
hexane and water sometimes formed. This would usually separate if 
allowed to sit undisturbed for a few minutes. If the emulsion did not 
separate, herbicide residues contained within the emulsion could have 
been discarded. There was an apparent contamination of the glassware 
used for water sample collection or extraction. Samples of deionized 



84 

water were extracted and they shewed concentrations of atrazine of 
approximately 1 /xg/L. All glassware was routinely washed with a 
commercial glassware detergent, triple rinsed with deionized water, and 
rinsed with acetone. The contamination levels appeared to be consistent 
in that water samples, in which atrazine was not expected, shewed a 
concentration of approximately 1 /xg/L. The source of the contamination 
was not determined. Several attempts were made to isolate and identify 
the source of the contamination, however these attempts were 
unsuccessful. To account for the contamination, the measured 
concentrations were reduced by 1 /xg/L. Thus no concentrations of less 
than 1 /xg/L are reported even though the sensitivity of the GC would 
allow detection down to the range of tenths of a /xg/L. 

The gas chromatograph performed well except that the N-P detector 
response had a tendency to drift. This was accounted for by running 
standards frequently and adjusting calculated concentrations as described 
in the methods section. 

The recoveries of atrazine and alachlor from spiked water samples 
were nearly identical and ranged from 68 to 104% with a mean of 79% and a 
standard deviation of 12 percent. The standard deviation would probably 
be reduced if these analyses had been performed by an experienced 
analyst. Experience contributes significantly to the ability to produce 
accurate and consistent results. 

The recoveries of atrazine and alachlor from spiked soil samples 
could not be determined from the data collected. Moisture contents of 
the soils used for spiking were not recorded. The spike levels were 
based on the moist weight of the soil. After spiking, the soil was air- 



85 

dried for 1 day in order to allow the solvent from the herbicide 
standards to evaporate prior to taking subsamples for analysis. The loss 
of water from the samples during this time resulted in an increased mass 
of pesticide per unit weight of wet soil. Since the initial moisture 
contents were unknown, it was not possible to calculate the expected 
concentrations in the air dried samples. Calculations based on the 
spiking levels of the moist soil resulted in apparent recoveries that 
exceeded 100 percent. It can be concluded from the data that recoveries 
from the soil were high. When the omissions in the soil recovery study 
were discovered, the facilities used to perform extractions and analyses 
were no longer available. 

All concentrations (water and soil) reported in the following 
sections are the analytically determined concentrations and have not been 
adjusted for expected recovery rates. 

6.3 Chemical Applications 
Applications of the herbicides and bromide tracer were monitored to 
determine the concentration of the chemicals in the application water and 
the volume of water applied. Concentrations of the chemicals were 
multiplied by the volume of water collected and divided by the opening 
area of the collectors in order to determine the application rate on a 
mass per unit area basis. Table 6.2 summarizes the measured chemical 
application data. The intended application rates were presented in Table 
5.3. 

The application of atrazine and alachlor to the designated 
application area was not uniform as can be seen in Figures 6.1 and 6.2. 



86 




a) Solution Concentration 



08-09 




b) Application Rate 



Figure 6.1. Uniformity of atrazine application, a) concentration in 
application solution (vertical bars show sampling 
locations) , b) application rate (vertical bars show 
location of monitoring wells) 



37 




08-09 




b) Application Rate 



Figure 6.2. Uniformity of alachlor application, a) concentration in 
application solution (vertical bars show sampling 
locations) , b) application rate (vertical bars show 
location of monitoring wells) 



88 

The application solution was used up just as the sprayer crossed over 
well 10-09 (6.1 m short of intended end of application) causing the rapid 
decrease in application rate at that point. Overall, the application 
showed considerable variability that is probably representative of 
chemical applications in general. Similar nonuniformity was observed 
during the second application of the bromide tracer as shown in Figure 
6.3. The samples taken during the first application of bromide were lost 
due to inadvertent freezing of the samples and subsequent breakage of the 
glass sample containers. 



Table 6.2 Chemical application results. 



Date of 
Application 


Chemical 


Depth of 
Water (cm) 1 


Concentration 
(mg/L) 1 


Application 
Rate (kg/ha) 1 


11/12/86 


Atrazine 


0.70 


51.5 


3.41 




(3D 2 


(28) 


(39) 


11/12/86 


Alachlor 


0.70 


11.1 


0.72 






(31) 


(34) 


(36) 


11/17/86 


Bromide 


0.86 


_3 


-3 




(from KBr) 


(20) 






4/27/87 


Bromide 


0.74 


111.0 


7.47 




(from KBr) 


(20) 


(87) 


(75) 



1 Mean of observations 
2 Coefficient of variation (%) 
3 Samples lost prior to analysis. 



The nonuniformity of application demonstrates one of the problems 
associated with prediction of chemical transport within the soil profile. 
In addition to recognized variability of soil properties, there is the 
(compounding influence of variability in application rates over a field. 
The concentrations of alachlor measured in the application samples were 



89 




a) Solution Concentration 




b) Application Rate 



Figure 6.3. Uniformity of bromide application on 4/27/87, 

a) concentration in application solution (vertical bars 
show sampling locations) , b) application rate (vertical 
bars show location of monitoring wells) 



90 

significantly lower than for atrazine although they were applied at the 
same rate (4.9 kg/ha) . Figure 6.4 shows the relationship between 
measured concentrations of alachlor and atrazine in the sample collection 
containers. The concentration of alachlor in a container is strongly 
correlated to the concentration of atrazine. The alachlor concentrations 
were, in general, only about 24% of the atrazine concentrations. 
Soil sample data taken six days after application (Table 6.3) suggests 
that more alachlor reached the soil than indicated by the application 
rate shown in Table 6.2. Possible explanations for the concentrations of 
alachlor in the application samples include errors in formulating the 
application solution, adsorption to the plastic cup, and volatilization. 



APPLICATION SOLUTION CONCENTRATIONS 

Y = -1.10 + 0.24 * X * 
CORR. COEF. = .89 



a 



15- 



u 

glOH 

K 
o 

u 

< o 



* ** 







1 1 1 I 1 1 1 1 I ' ' I I I I I I I I I I I I I I I I , 



'I I I I I I- 



10 20 30 40 50 60 70 
ATRAZINE CONC. (mg/L) 



Figure 6.4. Comparison of alachlor and atrazine concentrations in 
application samples. 



91 



Alachlor is more volatile than atrazine. However, these samples were 
exposed for less than 10 minutes after the application (while being 
collected, sealed in glass containers, and placed on ice) . 

6.4 Chemicals in the Unsaturated Zone 
6.4.1 Atrazine and Alachlor 

Samples of the top 5 cm of soil were taken from 14 locations within 
the application area on 11/18/86, which was six days after application. 
Five cm of irrigation water had been applied prior to this sampling. 
Table 6.3 compares the measured application rates of atrazine and 
alachlor shown in Table 6.2 with the average measured soil 
concentrations on this date. A simple calculation reveals that if the 
actual application rate of alachlor was 0.72 kg/ha and was entirely 
contained within the top 5 cm of soil, the maximum possible concentration 
in the soil would be approximately 1 mg/kg (assuming that the soil bulk 
density equals 1.45 g/cm 3 ) . The 3.41 mg/kg measured average value could 



Table 6.3 Measured application rates and 
soil surface concentrations 
of atrazine and alachlor. 



Application 
rate (kg/ha) 



Atrazine 
3.41 1 
(39) 2 



Alachlor 
0.72 



(36) 



Soil 

Concentration 
(mg/kg) 



0.29 



(68) 



3.26 



(58) 



Mean 



^Coefficient of variation (%) 



92 

only be achieved if the actual application rate were at least 2.5 kg/ha. 
If the planned application rate of 4.9 Jog/ha were contained within the 
top 5 cm of soil, the expected concentration in the soil would be 6.75 
mg/kg. The observed soil concentrations suggest that the alachlor in the 
application samples either volatilized rapidly or adsorbed to the 
collection containers prior to transfer into glass containers. If 
volatilization from the sample containers was the reason for the low 
concentrations, it would be expected that there would also have been 
significant volatilization from the grass foliage and soil surface prior 
to irrigating the site and moving the chemical off of the foliage and 
into the soil profile. 

The low concentration of atrazine within the top 5 cm of soil is 
probably due to movement below this depth with percolating water before 
collection of these samples. Soil samples collected six days later 
(11/24/86) showed considerably more atrazine in the zone from 5 to 15 cm 
than from to 5 cm. Samples from the 5-15 cm depth range were not 
collected on November 18. 

Figures 6.5 and 6.6 show the relationship between the measured soil 
concentration of atrazine in the top 5 cm of soil on 11/18/86 and the 
concentration of the application solution and the application rate, 
respectively. Figures 6.7 and 6.8 show the same relationships for 
alachlor. Figure 6.5 shows that there was a poor correlation between the 
concentration of the application solution and the concentration in the 
upper 5 cm of soil measured at the same locations 6 days after 
application. Figure 6.6 shows that there was a higher correlation 
between application rate and the soil concentration than there was for 



1.0 -, 



0.8 - 



\ 

a 



0.6 H 



O 0.2 H 
CO 



0.0 



SOIL CONC. VS. APPLICATION CONC. 



ATRAZINE APPLIED 11/12/86 
SOIL CONC. MEAS. 11/18/86 



93 



CORR. COEF. = .28 



1 1 1 1 i 11 1 1 i ' ' ' ■ i ii ii i i i ii i i i i i [ i i i i I i , 
10 20 30 40 50 60 70 

ATRAZINE CONC. (mg/L) 



Figure 6.5. Correlation between atrazine concentrations in the top 
5 cm of soil and application solution concentrations. 



1.0 



0.8 - 



tt) 

\ 

0.6 

6 



o 0.2 H 
in 



0.0 



SOIL CONC. VS. APPLICATION RATE 



ATRAZINE APPLIED 11/12/86 
SOIL CONC. MEAS. 11/18/86 



CORR. COEF. = .77 







1 ' 11 i 1 1 1 1 I i ' i i I i i i i i i i i i I i i i i i , , 
1 2 3 4 5 6 

ATRAZINE APPL. RATE (kg/ha) 



Figure 6.6. Correlation between atrazine concentrations in the top 
5 cm of soil and application rate. 



6.0 

5.0 
bo 

A 4.0 



SOIL CONC. VS. APPLICATION CONC. 



94 



o 
u 



3.0 - 



2.0 - 



§ 1.0 



0.0 



CORR. COEF. = .27 



ALACHLOR APPLIED 11 
SOIL CONC. MEAS 



D 11/12/ 
• ll/18/€ 



86 
86 







' ' ' 1 1 1 — 1 — 1 — 1 — ' — i — [ — i — i — f — , — , 
5 10 15 20 

ALACHLOR APPL. CONC. (mg/L) 



Figure 6.7. Correlation between alachlor concentrations in the top 
5 cm of soil and application solution concentrations. 



6.0 -, 

be 

a 



u 

o 
u 



4.0 - 



3.0 



2.0 - 



S i-o - 



0.0 - 



SOIL CONC. VS. APPLICATION RATE 



CORR mPTT - n ALACHLOR APPLIED 11/12/86 
TORR COEF 31 S0[L C0NC , MRAS , , /, 



ALACHLOR APPL. RATE (kg/ha) 

Figure 6.8. Correlation between alachlor concentrations in the top 
5 cm of soil and application rate. 



95 

application solution concentration and soil concentration. Figures 6.7 
and 6.8 show that there was essentially no correlation between either 
alachlor concentration in the application solution or application rate 
and the concentration of alachlor in the top 5 cm of soil. 

Additional soil samples were collected four more times during the 
study. In order to reduce the total number of samples collected, the 
number of depths sampled was increased and the number of sampling 
locations was decreased with each successive sampling date. Alachlor was 
not observed to move below a depth of 36 cm in the soil. No trace of 
alachlor was detected in the shallow groundwater samples. Consequently, 
most of the results discussed in subsequent sections will be limited to 
atrazine. Results of soil sample analyses for atrazine and alachlor 
residues are presented in Tables 6.4 and 6.5, respectively. Table 6.4 
shows that the sampling strategy to increase the depth of sampling with 
time since application resulted in missing the passage of the initial 
atrazine front past any point. Samples from a depth of 178-188 cm on 
12/22/87 already showed atrazine residues. Tables 6.4 and 6.5 include a 
calculation of the total mass of atrazine and alachlor in the soil 
profile beneath the application area, respectively. For these 
calculations, the concentration was assumed to vary linearly between 
sample points. The concentration of atrazine within the profile appears 
to increase with time for the first 4 sampling dates due to the way in 
which the samples were collected ( increasing depth of sampling over 
time. The maximum mass of atrazine within the soil profile was 
calculated to be 70 g on 2/9/87 which is approximately 47% of the 
intended application of 150 g, or 62% of the measured application 



Table 6.4. Mean concentrations of atrazine (mg/kg) in soil samples. 



Sample Depth 
(cm) 


11/18/86 


11/24/86 


Sample Date 
12/22/86 


2/9/87 


3/16/87 


5/25/87 


0-5 


0.29 1 
(68) 3 


0.23 
(61) 


NS 2 


NS 


0.06 
(42) 


NS 


0-15 


NS 


NS 


NS 


0.16 
(53) 


0.15 
(29) 


0.04 
(15) 


5-15 


NS 


0.55 
(47) 


NS 


NS 


NS 


NS 


25-36 


NS 


NS 


0.11 
(45) 


0.23 
(61) 


0.13 
(51) 


0.07 
(20) 


41-51 


NS 


NS 


NS 


0.15 
(30) 


0.08 
(71) 


NS 


56-66 


NS 


NS 


0.02 
(45) 


0.09 
(49) 


0.05 
(65) 


0.04 
(78) 


71-81 


NS 


NS 


NS 


0.05 
(53) 


NS 


NS 


86-97 


NS 


NS 


0.02 
(63) 


0.04 
(40) 


0.03 
(94) 


0.03 
(61) 


117-127 


NS 


NS 


NS 


0.03 
(58) 


0.03 
(125) 


NS 


132-142 


NS 


NS 


0.02 
(49) 


NS 


NS 


0.02 
(31) 


147-157 


NS 


NS 


01 
(77) 




u . uz 

(38) 


NS 


178-188 


NS 


NS 


0.01 
(64) 


NS 


0.04 
(67) 


0.01 
(38) 


Calculated 
Mass in 
Profile (g) 


7 


32 


40 


70 


58 


32 


-^Mean 


^No Sample 


•^Coefficient of variation 


(%) 





97 



Table 6.5. Mean concentrations of alachlor (mg/kg) in soil samples. 



Sample Depth 
(cm) 


11/18/86 


11/24/86 


Sample Date 
12/22/86 


2/9/87 


3/16/87 


5/25/87 


A— R 
U O 






no 


HQ 
JNO 


1 A"7 


XTC 




(58) 


(59) 






(55) 






XTC 


No 


XTC 


U. /o 


U • Do 


0. 12 










(104) 


(23) 


(65) 


5-15 


NS 


0.17 


NS 


NS 


NS 


NS 






(96) 










25-36 


NS 


NS 


0.00 


0.00 


0.02 


0.01 












(224) 


(141) 


Calculated 


79 


84 





53 


69 


13 



Mass in 
Profile (g) 

^Mean 2 No Sample ^Coefficient of variation (%) 

rate of 3.41 tog/ha over an application area of 334 m 2 (114 g) . The 
maximum mass of alachlor within the soil profile was calculated to be 79 
g on 11/18/86. This represents 53% of the intended application rate, and 
in excess of 300% of the calculated application rate based on the 
collection of the application solution as described above. 

The complete data set of soil sample concentrations can be provided 
upon request. Refer to Appendix B for information on how to request the 
data and a sample of the data set. 

Approximately 30 samples from the soil solution samplers were 
extracted and analyzed for residues of atrazine and alachlor. The 
samples from sampler 09N-2 showed a pulse of atrazine moving past the 61 
cm depth. The peak concentration of atrazine in samples from this 
sampler was approximately 0.35 mg/L and this peak occurred on Nov. 24, 



98 

1986, which was 16 days after application. A plot of the atrazine and 
bromide concentrations, from these samples, as a function of the total 
depth of water applied since application of each chemical is presented in 
Figure 6.9. Since the bromide was applied 5 days after the atrazine, the 
concentrations were plotted as a function of the total water applied 
since each chemical was applied to provide a common base for comparison. 
From this figure it can be seen that approximately 2.4 times as much 
water had to be applied to move the peak concentration of atrazine past 
the 61 cm depth as was required to move the bromide peak past the same 
point. This can be interpreted as an approximation of the retardation 
factor of atrazine in this soil (assuming that the bromide 




Figure 6.9. Bromide and atrazine concentrations in solution 

sampler 09N-2 as a function of total water applied 
since application. 



99 

tracer is nonadsorbed) . The retardation factor can be written as 

R = 1 + (Kd) (Pb) 
%c 

where R = retardation factor, dimensionless , = partition coefficient, 
Pb = soil bulk density, and e fc = soil-water content at field capacity 
(Dean et. al, 1984). The retardation factor indicates the velocity of 
water or a nonadsorbed chemical relative to the velocity of an adsorbed 
chemical. Thus a retardation factor of 2 would indicate that the 
adsorbed chemical would move at 1/2 of the velocity of a nonadsorbed 
species. 

Using the soil properties listed in Table 5.1 and a normalized 
partition coefficient for atrazine of 163 cm 3 /g) / a weighted average 
retardation factor for the top 61 cm of the soil profile was calculated 
to be approximately 6. The properties listed for the top 13 cm yield a 
retardation factor of 14 and the properties of the layer from 51-62 cm 
give a retardation factor of 2. 

Alachlor was not detected in any of the extracted soil solution 
samples. 

6.4.2 Bromide 

The concentrations of bromide (Br~) in the 61, 122, and 183 cm deep 
solution samplers following the first application of bromide to the site 
are shown in Figures 6.10-6.12, respectively. The first application of 
bromide was made 5 days after the application of atrazine and alachlor. 
The date of the application of the herbicides to the application site 
(11/12/86) is used as the benchmark time to which all observations and 



100 

results are referenced. These figures demonstrate the observed 
variability between sampling locations. Figure 6.13 shows the average 
concentration, from the 6 sampling locations, of bromide at each sample 
depth following the first bromide application. Figures 6.14-6.16 show 
the concentrations of bromide following the second application in the 
61, 123, and 183 cm solution samplers, respectively. Again, the 
variability between sampling locations is evident. Figure 6.17 shows the 
average concentration of bromide at each sample depth following the 
second bromide application. 

Tables 6.6 and 6.7 present the mean concentrations and coefficients 
of variation between sampling locations of bromide at each sampling depth 
on selected sampling dates following the first and second applications of 
bromide, respectively. These tables show that the coefficient of 
variation between sampling locations ranges from 23 to over 200%. 

The results of all analyses for tracers and herbicides in both soil 
and water samples can be provided on magnetic media. Refer to Appendix B 
for information on how to request this data and a sample of the data 
sets. 

6.4.3 Nitrate 

The entire field was fertilized on 4/16/87 at a rate of 560 kg/ha of 
5-10-15. Concentrations of nitrate (as N0 3 ) moving through the vadose 
zone following the fertilizer application exhibited extreme variability. 
There were virtually no detections of nitrate due to the fertilizer 
application in the samplers located at a 61 cm depth. Two samplers 
located at a depth of 122 cm shewed a response to the fertilizer 



^ 8-] 
-J : 

*\ 7 z Station 



° ° ° o ° 08N-2 

6- » OBS-2 

I ooooo 09N— 2 




DAYS SINCE 11/12/86 



Figure 6.10. Bromide concentration for six sairpling locations 
following the first application at a 61 cm depth. 




DAYS SINCE 11/12/86 



Figure 6.11. Bromide concentration for six sampling locations 
following the first application at a 122 cm depth. 



Station 

08S-6 

oojoo 08N— 6 
° o ° o ° 09S-6 
» .a 09N-6 
» » » o o 10S— 6 
10S-6 




DAYS SINCE 11/12/86 



Figure 6.12. Bromide concentration for six sampling locations 
following the first application at a 183 cm depth. 





8q 




7^ 














5{ 


U 




ON 


4i 


u 


3^ 


w 




Q 


2^ 


i— i 




2 


li 


o 


as 






0- 



AVERAGE CONC. AT EACH SAMPLING DEPTH 

SOLUTION SAMPLES 



61 cm 
122 cm 
183 cm 



a 

o 
3 

03 




i I i i i I | I 1 'r 
20 30 40 50 60 70 80 

DAYS SINCE 11/12/86 



Figure 6.13. Average bromide concentration for the three sampling 
depths following the first application. 



103 



25-, 



Station 

08S-2 

oopoo 08N-2 

a a a o a 09S — 2 

• »_? 09N-2 
» o » o » 10S— 2 
ION— 2 




170 180 190 200 

DAYS SINCE 11/12/86 



210 



Figure 6.14. Bromide concentration for six sampling locations 
following the second application at a 61 cm depth. 



25n 




Station 

08S-4 

oopoo 08N— 4 
□_o_o_p_o 09S— 4 
* *_? 09N-4 

» 9 9 9 10S — 4 

10N-4 



160 



170 180 190 

DAYS SINCE 11/12/86 



T 1 1 1 1 

200 210 



Figure 6.15. Bromide concentration for six sampling locations 

following the second application at a 122 cm depth. 



104 



25 h 



Station 

08S-6 

o <^o oo OBN— 6 
° ° o ° o 09S — 6 
* *_? 09N-6 
» » » » » 10S— 6 
ION— 6 




i ' ' — i — i — i — i — i |— i — i — i — i — i 
190 200 210 



DAYS SINCE 11/12/86 



Figure 6.16. Bromide concentration for six sampling locations 

following the second application at a 183 cm depth. 





25 


»— ] 

\ 










20 : 














O 




o 




w 




MID 




5^ 


o 








PQ 


o- 



AVERAGE CONC. AT EACH SAMPLING DEPTH 



SOLUTION SAMPLES 

61 cm 

A *_5 122 cm 
•_• 183 cm 




160 



~i I — i — i — i "~i P i — i — i — i — | 
190 200 210 

DAYS SINCE 11/12/86 



Figure 6.17. Average bromide concentration for the three sampling 
depths following the second application. 



105 



Table 6.6. Mean concentration of bromide (mg/L) at each sampling depth 
following first application 1 . 



Sample Date 


61 


Sample Depth (cm) 
122 


183 




11/17/86 


0.03 Z (115) J 


00 


\ v i 


0.00 


(* 4 ) 


11/18/86 


0.05 


(114) 


0.10 


(82) 


0.10 


(*) 


11/24/86 


1.92 


(106) 


0.72 


(65) 


0.05 


(135) 


12/01/86 


1.87 


(50) 


-J • -J -J 


(34} 


0.31 


(146) 


12/08/86 


0.90 


(41) 


3.28 


(58) 


1.48 


(54) 


12/15/86 


0.39 


(41) 


3 07 




2.10 


(34) 


12/22/86 


0.25 


(45) 


1.42 


(53) 


1.58 


(23) 


12/29/86 


0.22 


(63) 


1.15 


(58) 


1.37 


(65) 


01/05/87 


0.08 


(225) 


0.85 


(70) 


0.91 


(59) 


01/12/87 


0.11 


(96) 


0.75 


(71) 


0.75 


(56) 


01/19/87 


0.04 


(65) 


0.68 


(89) 


0.68 


(73) 


01/26/87 


0.04 


(44) 


0.65 


(58) 


0.29 


(156) 


02/02/87 


0.09 


(156) 


0.52 


(78) 


0.51 


(60) 


02/09/87 


0.12 


(191) 


0.45 


(65) 


0.67 


(81) 



Bromide applied 11/17/86 2 Mean 
Coefficient of variation (%) 4 One sample 



106 



Table 6.7. Mean concentration of bromide (mg/L) at each sampling depth 
following second application 1 . 



Sample Date 


61 


Sample Depth (cm) 
122 


183 




04/27/87 


0.06^(32) J 


fl 9R 




0.31 


(*4) 


04/28/87 


2.06 


(186) 


n in 




0.11 


(*) 


04/29/87 


6.28 


(133) 






0.56 


(*) 


04/30/87 


7.81 


(112) 






0.95 


(93) 


05/01/87 


7.92 


(82) 


n qq 


\ ^ 


1.47 


(127) 


05/02/87 


5.56 


(116) 


A TR 




2.02 


(114) 


05/03/87 


5.54 


(75) 


9.02 


(77) 


2.96 


(112) 


05/05/87 


3.47 


(68) 


6.79 


(51) 


4.62 


(100) 


05/08/87 


1.58 


(104) 


5.44 


(49) 


6.97 


(128) 


05/11/87 


1.14 


(96) 


7.78 


(76) 


4.99 


(82) 


05/13/87 


0.79 


(99) 


6.11 


(*) 


1.91 


(56) 


05/25/87 


0.51 


(158) 


1.32 


(56) 


3.04 


(113) 


06/01/87 


0.10 


(72) 


1.05 


(63) 


2.20 


(130) 



-■■Bromide applied 04/27/87 2 Mean 
3 Coefficient of variation (%) 4 0ne sample 



application. The maximum concentration of nitrate in these samplers was 
5.6 mg/L on 5/03/87. Three samplers located at a depth of 183 cm 
responded to the application of fertilizer. The maximum concentration of 
nitrate in these samplers was 8.1 mg/L on 5/08/87. Concentrations of 
nitrate in the soil solution were not as high as observed concentrations 
within the groundwater over much of the study site (see Section 6.5.3 for 
nitrate concentrations in the saturated zone) . This is due to the fact 
that the tractor-mounted broadcast spreader was unable to get close to 



107 

the instrumented application area due to the presence of the vacuum and 
sample tubing and the instrument trailer. As a result, the application 
rate of fertilizer in this area was reduced. Refer to Appendix B for 
information on the availability of the data set containing results of all 
nitrate analyses. 

6.5 Chemicals in the Saturated Zone 

6.5.1 Bromide 

There was no shallow groundwater present on the site between the 
date of herbicide application (11/12/1986) and Dec. 15, 1986. Bromide 
was observed in the groundwater on this first well sampling date at a 
concentration of 0.8 mg/L in well 08-09 (see Figure 4.1 for well 
locations) which is within the application area. The continued detection 
of bromide in the groundwater following the first application was 
sporadic, and the maximum concentration of 3.9 mg/L was detected in well 
09-11 on January 12, 1987. The maximum concentration of bromide in the 
groundwater following the second application (4/27/87) was 5.2 mg/L on 
May 13th, in well 08-09. Figures 6.18-6.23 show the concentration of 
bromide within the groundwater over six sampling periods following the 
first application of bromide. Figures 6.24-6.29 show the concentration 
of bromide within the groundwater for six sampling periods following the 
second application of bromide. 



108 



Date: 12/15/86 

08-09 




Figure 6.18. Bromide concentration in the groundwater on 12/15/86. 
Vertical bars indicate sampling locations. 



Date: 12/22/86 




Figure 6.19. Bromide concentration in the groundwater on 12/22/86. 
Vertical bars indicate sampling locations. 



109 



09-11 Date: 12/29/86 




Figure 6.20. Bromide concentration in the groundwater on 12/29/86. 
Vertical bars indicate sampling locations. 



Date: 01/05/87 




Figure 6.21. Bromide concentration in the groundwater on 1/05/87. 
Vertical bars indicate sampling locations. 



110 



09-1 1 



Date: 01/12/87 




<Vf; no 



190 



Figure 6.22. Bromide concentration in the groundwater on 1/12/87. 
Vertical bars indicate sampling locations. 



Date: 01/19/87 



09-1 1 




1 3n 

0| ST ANCE - ' 50 



19 57 |5TA r# 



Figure 6.23. Bromide concentration in the groundwater on 1/19/87. 
Vertical bars indicate sampling locations. 



Ill 



Date: 05/05/87 




Figure 6.24. Bromide concentration in the groundwater on 5/05/87. 
Vertical bars indicate sampling locations. 



Date: 05/08/87 

08-09 




Figure 6.25. Bromide concentration in the groundwater on 5/08/87. 
Vertical bars indicate sampling locations. 



112 



08-09 




Figure 6.26. Bromide concentration in the groundwater on 5/13/87. 
Vertical bars indicate sampling locations. 



Date: 05/18/87 

08-13 




Figure 6.27. Bromide concentration in the groundwater on 5/18/87. 
Vertical bars indicate sampling locations. 



113 



10-09 




Figure 6.28. Bromide concentration in the groundwater on 5/25/87. 
Vertical bars indicate sampling locations. 



09-1 1 




Figure 6.29. Bromide concentration in the groundwater on 6/01/87. 
Vertical bars indicate sampling locations. 



114 

6.5.2 Atrazine 

Atrazine first appeared in the groundwater beneath the application 
area in the samples of Jan. 21. By this date, approximately 45 cm of 
water had been applied to the site as rainfall and irrigation since 
application of the herbicides. 

Groundwater samples from Jan. 21 through June 6 show what appear to 
be distinct pulses of atrazine reaching the saturated zone and moving 
downslope with the saturated flow. These pulses are characterized by low 
initial concentrations in the groundwater followed by the appearance of 
atrazine beneath the application area and movement within the groundwater 
as the water flows downgradient to the west. Figures 6.30 through 6.35 
show the concentration of atrazine (in /ug/L) in the groundwater at 
weekly intervals for a five week period beginning on Feb. 23, 1987. A 
vertical bar in these figures indicates that a well sample was collected 
and analyzed from that location. The transport of atrazine with the 
saturated flow is clearly visible and atrazine moved a distance of 
approximately 85 m in a period of 21 days. 

Figure 6.36 shows the general directions of flow based upon the 
slope of the water table on May 8, 1987. Hie slope of the water table on 
other sampling dates shows that flow would occur in the same general 
directions. In Figures 6.30 through 6.32 it is observed that atrazine 
has moved downslope from well 09-11 to 09-14. The average gradient of 
the water table between these wells during this period was approximately 
0.05 m/m. The hydraulic conductivity of the soil was not measured during 
this study. Assuming a hydraulic conductivity of 14.4 m/day (60 cm/hr) 
at a depth of approximately 3 m (Carlisle et al., 1978) and an effective 



115 



09-11 




Figure 6.30. Atrazine concentration in the groundwater on 2/23/87. 
Vertical bars indicate sampling locations. 




Figure 6.31. Atrazine concentration in the groundwater on 3/02/87. 
Vertical bars indicate sampling locations. 



116 



09-11 
08-09 



Date: 03/09/87 



09-14 




190 



Figure 6.32. Atrazine concentration in the groundwater on 3/09/87. 
Vertical bars indicate sampling locations. 



Date: 03/16/87 




Figure 6.33. Atrazine concentration in the groundwater on 3/16/87. 
Vertical bars indicate sampling locations. 



117 



09-11 Date : 03/23/87 




Figure 6.34. Atrazine concentration in the groundwater on 3/23/87. 
Vertical bars indicate sampling locations. 



Date: 03/31/87 




Figure 6.35. Atrazine concentration in the groundwater on 3/31/87. 
Vertical bars indicate sampling locations. 



118 

Water Table Elevation (m) 5/08/87 




Distance (m) 

Figure 6.36. Contour plot of water table elevation on 5/08/87 
showing direction of flow. 

saturated porosity of 0.36 (Table 5.1) , it is possible to calculate an 
estimated water velocity between the wells. The average pore water 
velocity is calculated as 

v = k * s / p e 6>1 
where v = average pore water velocity, m/day 

k = saturated hydraulic conductivity, m/day 
s = hydraulic gradient, m/m 
Pe = effective saturated porosity, cm 3 /cm 3 

Using Equation 6.1 and the values of the variables described above, 
a pore-water velocity of 2 m/day was calculated. The distance between 
wells 09-11 and 09-14 is 36 m. Based on the calculated pore-water 
velocity, water should move the distance between the wells in 18 days. 
Atrazine was observed to have moved the distance between the wells in 14 



119 

days. This would suggest that atrazine was moving faster than the 
velocity of the water. In Figures 6.32 and 6.33 it is observed that 
atrazine moved the distance from well 09-14 to well I£W in a period of 7 
days. The water table gradient between these wells was approximately 
0.7 m/m. The average pore water velocity for the area between these 
wells was calculated to be approximately 3 m/day. These wells are 
located a distance of 42 m apart. Based on the calculated water velocity, 
it should take 14 days for water to move the distance between the wells. 
Thus, using this method, atrazine would not be expected to have traversed 
the distance between these wells in 7 days. These simple calculations 
indicate that the conductivity value is probably too low. It is also 
possible that atrazine is moving along the top of the restricting layer 
in large pores or conduits which are not represented by the conductivity 
value used in the calculations. Without measured values of the saturated 
conductivity in the saturated zone, it is difficult to explain the rapid 
movement of atrazine within the groundwater. 

Since samples were collected on 7-day intervals and the wells are 
spaced 12 m apart, there is considerable error associated with estimates 
of the distance traveled between successive dates and the time required 
for atrazine to have moved a specific distance. There are also 
significant errors associated with assuming conductivity values for this 
site from general soil characterization data. The above examples 
illustrate that additional site specific data (hydraulic conductivity, 
effective porosity, etc.) are needed to adequately characterize the 
observed atrazine transport. 



120 

In Figure 6.37, the concentration of atrazine in one well is plotted 
along with the water table elevation in the well. There appears to be an 
inverse relationship between concentration and water table elevation or 
depth of the saturated zone. This well is 20 m downslope from the 
application area. Based on this figure, it is hypothesized that as the 
water table rises, the water around the well is primarily coming from 
percolation through the untreated soil above it which would cause the 
concentration of atrazine in the vicinity to decrease. During the period 
that the water table is receding, water which has infiltrated through the 
application area is flowing past the well, thus increasing the 
concentration . 




DAYS 



1 I ' ' I I I I I I I I T | "l I I | T- 

120 140 160 180 200 

SINCE 11/12/86 



co 

CV! 



Figure 6.37. Concentration of atrazine and water table elevation in 
well 09-11. 



121 

All concentrations of atrazine in the groundwater were below 0.1 
mg/L and in most cases were below 0.01 mg/L. The highest observed 
concentration was 0.09 mg/L and occurred on 5/5/87 in well 08-09 which is 
located within the application area. This concentration was observed 
after the water table had risen due to the application of approximately 
24 cm of irrigation water within a period of 8 days. There were high 
concentrations of atrazine in the groundwater when sampling was 
discontinued on 6/1/87. Thus, the data collected do not indicate how 
long atrazine residues persisted on the site. Figures D.1-D.23 in 
Appendix D show the concentration of atrazine in the groundwater on every 
sampling date from 1/19/87 through 6/01/87. 

6.5.3 Nitrate 

Background concentrations of nitrate (as NO3) in the groundwater 
were generally below 1 mg/L. Well LCW, which may respond to drainage 
from other parts of the research farm, often had background 
concentrations exceeding 15 mg/L. The study site was fertilized on April 
16, 1987, and was heavily irrigated beginning on April 27th. The water 
table on the site started to rise on May 1st. Concentrations of nitrate 
in the groundwater began increasing on May 3rd. Nitrate concentration 
levels in the groundwater were observed to be as high as 35 mg/L. 
Figures 6.38-6.45 show the nitrate concentrations in the groundwater on 
selected dates. 



122 



Date: 05/01/87 




Figure 6.38. Nitrate concentration in the groundwater on 5/01/87. 
Vertical bars indicate sampling locations. 



07-10 

07 _ 12 Date: 05/03/87 




Figure 6.39. Nitrate concentration in the groundwater on 5/03/87. 
Vertical bars indicate sampling locations. 



123 



08-10 



Date: 05/05/87 




Figure 6.40. Nitrate concentration in the groundwater on 5/05/87. 
Vertical bars indicate sanpling locations. 



Date: 05/08/87 



10-12 



07-14 




H 190 57 |STA Ntf 1 

Figure 6.41. Nitrate concentration in the groundwater on 5/08/87. 
Vertical bars indicate sanpling locations. 



124 



09-14 Date: 05/13/87 




Figure 6.42. Nitrate concentration in the groundwater on 5/13/87. 
Vertical bars indicate sampling locations. 



Date: 05/18/87 




Figure 6.43. Nitrate concentration in the groundwater on 5/18/87. 
Vertical bars indicate sampling locations. 



125 



Date: 05/25/87 




Figure 6.44. Nitrate concentration in the groundwater on 5/25/87. 
Vertical bars indicate sampling locations. 



Date: 06/01 /87 




Figure 6.45. Nitrate concentration in the groundwater on 6/01/87. 
Vertical bars indicate sampling locations. 



126 

6.5.4 Chloride 

In order to further characterize the flow velocities within the 
groundwater, 10 L of solution containing 23.8 g/L of chloride (50 g/L of 
KC1) was poured into well 07-09 on 5/1/87. The chloride was added to the 
groundwater at a time when nearly continuous percolation was occurring 
due to irrigation, and the water table was beginning to rise. Chloride 
concentrations in wells downslope from well 07-09 showed chloride levels 
exceeding 80 mg/L following the chloride application. 

Figures 6.46-6.53 show the chloride concentrations within the 
saturated zone on selected dates. In Figure 6.46 the background 
concentration of chloride on May 1, 1987 can be seen to be on the order 
of 1 mg/L or less. Figure 6.47 shows that chloride from other sources 
(KC1 in the fertilizer) is also entering the groundwater. This is most 
clearly demonstrated by the high concentration of chloride in well 11-10 
which is on the opposite side of the study site from well 07-09 which is 
where the chloride was introduced into the groundwater. Figures 6.46- 
6.53 show the general movement of chloride in the groundwater over a 
period of 29 days. 

Discussions presented in Section 6.5.6 will show that additional 
characterization of soil properties near the restricting layer is needed 
before the observations of chemical movement within the groundwater can 
be used to describe flow patterns and velocities within the groundwater. 
No calculations of flow velocities were performed with the chloride data 
and no conclusions were drawn from it for presentation here. 



127 



Date: 05/01/87 

07-09 




Figure 6.46. Chloride concentration in the groundwater on 5/01/87. 
Vertical bars indicate sampling locations. 



07-10 

Date: 05/03/87 




Figure 6.47. Chloride concentration in the groundwater on 5/03/87. 
Vertical bars indicate sampling locations. 



128 



Date: 05/05/87 




Figure 6.48. Chloride concentration in the groundwater on 5/05/87. 
Vertical bars indicate sampling locations. 



Dcte: 05/08/87 

08-12 




Figure 6.49. Qiloride concentration in the groundwater on 5/08/87. 
Vertical bars indicate sampling locations. 



129 




Figure 6.50. Chloride concentration in the groundwater on 5/13/87. 
Vertical bars indicate sampling locations. 




Figure 6.51. Chloride concentration in the groundwater on 5/18/87. 
Vertical bars indicate sampling locations. 



130 



08-1 1 




Figure 6.52. Chloride concentration in the groundwater on 5/25/87. 
Vertical bars indicate sampling locations. 



Date: 06/01 /87 




Figure 6.53. Chloride concentration in the groundwater on 6/01/87. 
Vertical bars indicate sampling locations. 



131 

6.5.5 Total mass of chemicals within the saturated zone 

In addition to observations of the concentrations of the chemicals 
within the groundwater, calculations of the total mass of chemicals 
within the saturated zone were performed. An inverse distance weighting 
method was used for each observation period to generate a square grid of 
values over the site. One potential option for gridding data was kriging 
which has been shown to be more accurate than the inverse distance method 
(Golden Software, 1987) . When kriging was used, however, it was observed 
that negative concentration values were generated at several intermediate 
grid points. With the inverse distance method, negative concentrations 
were not a problem. The data from observation wells spaced 12 m by 12 m 
apart were gridded to produce data on a 3m by 3m spacing. This was done 
initially for illustrative purposes. 

The data from the wells were gridded in two ways. First, only the 
wells from which samples were obtained and analyzed were included in the 
input data set for the gridding process. Second, all wells were included 
by assigning the concentration in wells without samples a value of zero. 
The second method was used as an estimate of the minimum mass of a 
chemical within the saturated zone. The three-dimensional views of the 
concentrations were judged to better convey to the viewer the values at 
individual wells when the second method was used as compared to the first 
method. 

Using the concentrations and groundwater depths computed on a 3 m 
square grid, calculations of water storage and chemical mass storage 
within the groundwater were performed. These calculations are described 
in more detail in Section 6.5.6. 



132 

The results of these calculations for the two methods of gridding 
are presented in Table 6.8. Note that the mass of chloride within the 
groundwater greatly exceeds the amount applied. Significant 
concentrations of chloride were apparently contributed from other sources 
such as fertilizer and irrigation water. There are notable differences 
in the total mass of chemicals stored as calculated from the two methods 
of gridding the observed data. There is no evidence to suggest that one 
method is more correct than the other, therefore the values calculated 
based on the two methods can be considered as the extremes of the 
possible mass storage in the groundwater with the actual value probably 
lying somewhere between the extremes. Figure 6.54 shows the mass of 
atrazine in the saturated zone (based on the second gridding method) 
during the entire period of observation. What is not shown by Figure 
6.54 is the mass of atrazine which may have been transported off-site 
during the period of observation. In order to estimate the transport of 
atrazine from the site, it is necessary to first calculate the flux of 
water from the site. The next section describes the methods used to try 
to estimate the water and chemical fluxes on the experimental site. 



Table 6.8 Total mass of chemicals in the saturated zone 



133 



Chemical 


Date of 
Maximum 


Maximum Mass 
Method l 1 


in Saturated Zone (g) 
Method 2 1 


Atrazine 


5/18/87 


8 


4 






(5) 


(3) 


Bromide 


5/13/87 


308 


198 






(54) 


(35) 


Nitrate 


5/05/87 


2916 


1873 






(23) 


(15) 


Chloride 


5/13/87 


16367 


11812 






(6882) 


(4967) 



-■-Method 1: using only measured values in gridding procedure 
Method 2: assuming that wells without observations had a concentration 
of zero 

2 Numbers in () are percent of total application 



6n 



ATRAZINE MASS IN WATER TABLE 

[FTT 




40 60 80 

DAYS SINCE 



i' i 1 1 1 i 1 1 

180 200 

11/12/86 



Figure 6.54. Total mass of atrazine stored in the saturated zone. 
Atrazine was applied to the site on 11/12/86. 



134 

6.5.6 Water Balance Calculations 

An attempt was made to calculate a water and chemical budget for the 
study site. The primary objective of this exercise was to calculate the 
mass flux of chemicals moving off the site. The mass of a chemical 
stored in the saturated zone at the end of the monitoring period plus the 
mass of chemical transported off-site should be comparable to the total 
mass of chemical leached from the root zone as predicted by the models. 

6.5.6.1 Nodes and subareas 

As noted in section 6.5.5, the data for water table elevation, 
restricting layer elevation, and chemical concentrations were gridded on 
a 3 m by 3 m spacing. Each grid point was considered to be a node for 
the calculations described below. The study area was divided into six 
subareas as shown in Figure 6.55. The subareas were numbered from the 
top of the slope. The chemical application area was entirely contained 
within subarea number 2 as shown in Figure 6.55. 

6.5.6.2 Nodal calculations 

The slope of the water table in both the x and y directions was 
calculated at every node within the site using centered-dif ference 
techniques, ie. along the y-axis the slope at node j is calculated as: 

slope(j) = [elev(j-l) - elev( j+1) ] / [2 * 6y] 6.1 
where slope(j) = slope of the water table in m/m, elev(j-l) and elev(j+l) 
= elevation of the water table at nodes j-1 and j+1, respectively, in m, 
and <Sy is the spacing between nodes in m. The gradient of the water 



135 




Distance (m) 
Figure 6.55. Subareas used in water balance calculations. 



table, as used in flux calculations, is the negative of the slope defined 
above. The flux of the water at any node was calculated as: 

flux(j) = -cond(j) * slope(j) 6.2 
where flux(j) = the water flux at node j in m/day, cond(j) = the 
saturated conductivity at node j in m/day, and slope(j) is as defined 
above. The volumetric flowrate of water through the area represented by 
node j was calculated as 

Q(j) = flux(j) * wtd(j) * 6y 6.3 
where Q(j) = volumetric flowrate past node j in m 3 /day, wtd(j) = 
groundwater depth at node j in m, and flux(j) and <5y are as defined 
above. The mass of chemical moving past a node was calculated as the 
product of the volumetric flowrate, Q, and the solution concentration of 
the chemical at that node. 

To compute the change in water storage between sampling periods, it 
was necessary to calculate the relationship between drained volume and 



136 

water table depth. The slope of the drained volume-water table depth 
relationship is called the drainable porosity. A program called DVOLWTD 
was used to calculate the drainable porosity. The DVOLWTD program is 
provided with the water table management model DRAINMDD (Skaggs, 1978) . 
The program uses the soil-water characteristic data for the profile and 
computes the volume of water drained as the water table falls. The soil 
water characteristic data input to DVOLWTD was obtained from Hook 
(1985) . Graphs of the drained volume-water table depth relationship and 
the soil-water characteristic curve are presented in Figures C.3 and C.4 
in Appendix C. The drained volume was calculated at every node based 
upon the depth to the water table at that node. 

The mass of chemical stored (sorted and in solution) in the volume 
represented by each node was calculated based upon the groundwater depth 
at the node, the concentration of the chemical in solution, and the 
partitioning coefficient for the chemical. The for atrazine used in 
this study is 163 cm 3 /g (Table 5.2) . The organic carbon content of the 
soil above the restricting layer was assumed to be 0.03% (Table 5.1) . 
Thus the Kq for atrazine in the lower soil zones is 0.049. The drained 
volume and chemical storage values calculated as described above were 
summed over each subarea. 

6.5.6.2 Boundary and subarea flux calculations 

The total volumetric flowrate of water was calculated for each 
boundary in a subarea. This was done by summing the flowrates calculated 
above for all the nodes along a boundary. The sum of the flowrates 
calculated for the four boundaries yields the net flowrate into or out of 



137 

the subarea on that date. The calculations described above were 
performed for all nodes and subareas for the 28 sampling dates in this 
study. 

6.5.6.3 Mass balances between observation dates 

The net flowrate between sampling dates for each subarea was 
calculated by averaging the net flowrates calculated on the two dates. 
The average net flowrate multiplied by the interval between the sampling 
dates resulted in an estimate of the total volume of water which had 
moved into or out of a subarea by saturated flow between the dates. The 
percolation volume during the interval between dates predicted by FRZM 
was used to calculate the volume of water added to each subarea due to 
rainfall or irrigation. The sum of the volume of water added or removed 
from a subarea by saturated flow and the volume added by percolation gave 
an estimate of the predicted change in water volume in the subarea 
between the sampling dates. 

The change in drained volume in a subarea between sampling dates was 
assumed to represent the actual change in water volume in that subarea. 
Thus, the predicted change in water volume could be compared to a 
"measured" value. 

Mass balances of chemicals within the saturated zone were done in a 
similar manner. Chemicals were assumed to be added to the saturated zone 
only in subarea number 2 which contained the application area. The mass 
of chemical leaching past the bottom of the soil profile (2.62 m) 
predicted by PRZM was multiplied by the area of the application area to 
calculate the total mass of chemical added to subarea number 2. 



138 

Prediction of mass transport across boundaries was performed as 
described above for water. The mass storage of chemical calculated for 
each subarea was compared to the predicted storage based on mass inflows 
and outflows to a subarea. 

6.5.6.4 Results 

The conductivity of the soil in the saturated zone was not measured. 
As an initial estimate, a conductivity of 14.4 m/day (60 cm/hr) was used 
(Carlisle et al . , 1978) . For periods when the water table was falling 
and no percolation was added, the predicted and observed change in water 
storage on 4 of the subareas agreed reasonably well. Subareas 3 and 6, 
however, would accumulate water when such accumulation was not observed. 
In order to avoid accumulation of water within subareas 4 and 6, a 
conductivity value on the order of 40 m/day was required along the 
boundary between subareas 4 and 5, and the boundary at the bottom 
(western edge) of area 6. This would suggest that the cross-sectional 
area of flow across these boundaries was not adequately characterized. 

With the higher conductivity values at the boundaries described 
above, the predicted and observed changes in water balances for the site 
were in reasonable agreement when no percolation was added. When 
percolation was predicted during the interval between sampling dates, 
however, the mass balances were not acceptable. The calculated change 
in storage in a subarea suggested that water accumulated within the 
subarea when measurements indicated that there was a net loss of water in 
that area. Whenever percolation was added, the predicted change in 



139 

storage would nearly always be substantially (order of magnitude) higher 
than the observed change. 

There are a number of possible reasons why the mass balances of 
water did not agree. One reason is that both FRZM and GLEAMS assume that 
all percolation through the soil profile occurs within one day. The 
actual movement of percolating water through the profile may be much 
slower than this and thus, may be sustained over significantly longer 
periods. Therefore, water which was predicted to have been added to the 
profile during the interval between sampling dates may actually have 
slowly entered the groundwater over several sampling intervals. It is 
possible that ERZM was overestimating the volume of percolation. GLEAMS, 
however, predicted nearly identical percolation volumes using an entirely 
different calculation procedure. This would suggest that the predicted 
percolation volumes are reasonable. 

Another source of error may be the assumption that the conductivity 
is constant over the depth of the groundwater. The conductivity may 
increase as the groundwater depth increases. There may also be large 
pores or conduits near the restricting layer which can move water much 
more rapidly than would be predicted using the methods described above. 
There is also significant error associated with the actual location of 
the restricting layer. The bottom of the wells were assumed to be placed 
on top of this layer. LXiring placement of the wells, the augered holes 
did not stop exactly on top of the restricting layer. The auger had to 
cut into the restricting layer in order visually determine (due to color 
and texture change) that the restricting layer had been reached. There 
could easily be variations in the depth of penetration in the restricting 



140 

layer of 5 cm. Since the depth of flow on this site was often less than 
10 cm, an error of +/~ 5 cm in the location of the restricting layer 
would significantly affect the calculated depth of flow. As noted above, 
the restricting layer was determined by visual observation of color and 
texture changes. The depth at which the vertical hydraulic conductivity 
of this layer becomes small enough to cause saturation may not correspond 
to the visually determined top of the layer. There may also be seme 
small channels on the surface of the restricting layer in which 
significant flow could occur that are not shown by the wells. This is 
demonstrated by the fact that conductivities along the two boundaries 
mentioned above had to be increased significantly above the values used 
elsewhere in order to move sufficient quantities of water across those 
boundaries. The potential water movement into the restricting layer was 
also not accounted for. 

There is much additional work to be done before an adequate water 
balance can be performed on this site. Additional characterization of 
soil properties in the saturated zone and improved estimates of the 
location of the restricting layer are needed. 

Since the water balance attempts were generally unsuccessful, 
chemical balances, although calculated, were not considered further. 

Two FORTRAN programs were written to perform the calculation 
described in the previous section. Program ANALYZE performs nodal and 
boundary flux calculations for each observation date and is presented in 
Figure C.l in Appendix C along with a sample output (Table C.l) . Program 
FLUX computes the water and chemical balances between sampling periods 



141 

and is presented in Figure C.2 in Appendix C along with a sample output 
(Table C.2) . 

6.5.7 Summary of observations of chemical transport 

The discussion of the observed concentrations of atrazine and the 
tracers has been descriptive in nature. This is due to the fact that 
additional data are required in order to quantitatively assess the 
results. 

Each of the applied chemicals (except for alachlor) was observed to 
move through the soil profile and enter the groundwater. However, 
insufficient data were collected from which the quantity or velocity of 
water percolating through the soil profile could be directly determined. 
Similarly, movement of chemicals within the groundwater was observed. 
There were insufficient data, however, to quantitatively assess the 
volume and velocity of flow within the saturated zone to acceptable 
levels. The tracer data were inconclusive due to the transient nature of 
the groundwater. Some of the wells would dry rapidly. The wells along 
the boundary between subareas 4 and 5 (Figure 55) would often be dry when 
wells on both sides of this boundary would have water in them. This 
would indicate that some saturated flow was occurring across this 
boundary in small channels in the restricting layer which did not 
coincide with the established well network ie. the wells did not 
intercept these channels. 



6.6 Model Results and Comparisons 



142 



The PRZM and GLEAMS models were run to simulate the time period from 
1/1/86 through 12/31/87. Simulation of the system for the ten months in 
1986 prior to application was done to minimize the effects of initial 
conditions. The simulated total mass flux of pesticides, bromide, and 
nitrates below the root zone during the sampling period are shown in 
Table 6.9. The mass flux of the chemicals to the bottom of the profile 
as simulated by PRZM is also included in Table 6.9. The models 
predicted that no additional movement of the applied chemicals below the 
root zone occurred after 6/1/87. 



Table 6.9 Simulated mass flux of chemicals. 



Chemical 


Simulated Mass Flux 


fq/ha) 


GLEAMS 1 


przm 1 


PRZM 2 


Atrazine 4 


2970 


2314 


1154 




(60. 6) 3 


(47.2) 


(23.6) 


Alachlor 5 


516 


165 


10 




(10.5) 


(3.4) 


(0.2) 


Bromide 6 


23604 


23718 


17116 




(87.4) 


(87.8) 


(63.4) 


Nitrate 6 


10657 


14551 


9376 




(57.3) 


(78.23) 


(50.4) 



■Bottom of root zone 4 Plant uptake coef . = 0.65 
Bottom of profile 5 Plant uptake coef. =0.52 

Percent of application 6 Plant uptake coef. =1.00 



GLEAMS predicted a higher percentage of herbicide leaching than did 
PRZM. One reason for the difference may be that GLEAMS defines 6 major 
soil layers (there is a 1 cm layer at the soil surface) and thus averages 



143 

soil properties such as the organic carbon content over larger depths 
relative to the smaller layers utilized by PRZM. The increased number of 
layers in PRZM allows the calculated layer properties to more nearly 
match the profile description input into the model. Another reason for 
the differences in model predictions may be the method by which pesticide 
transport is calculated in each model. In PRZM, a form of the 
advecticn-dispersion equation is solved (after water velocities have been 
determined separately) using finite-difference techniques to calculate 
pesticide transport. In GLEAMS, pesticide transport is calculated by 
sequentially moving the chemicals between layers based on water flux and 
the concentration of the chemical in each layer. After the chemicals 
have been added to or removed from a layer, the mass of the chemical is 
redistributed between the sorbed and solution phases based upon the 
partition coefficient in that layer. 

The predicted leaching of bromide by the two models was similar. 
Closer examination indicates that GLEAMS predicted higher fluxes of 
bromide for the first application, and PRZM. predicted higher fluxes for 
the second application. The overall effect was to predict nearly 
identical bromide fluxes. The nitrate leaching predicted by PRZM was 
higher than the GLEAMS prediction due to PRZM's higher predicted 
percolation volumes after the nitrate was applied as shown in Figure 
6.56. Overall, the percolation predicted by the models is very similar. 
A small adjustment in the leaf area index (LAI) in GLEAMS could result in 
closer agreement between the percolation predicted by the models. 

Figures 6.57-6.62 show the measured concentrations of atrazine in 
the soil profile and the corresponding predictions by PRZM. Figures 



144 

6.63-6.68 present similar data for alachlor. Results from GLEAMS are not 
included in these figures due to the fact that GLEAMS only outputs 
results for days on which there was sufficient rainfall to cause 
percolation beneath the root zone. There were no GLEAMS results 
available for most of the days shown in Figures 6.57-6.68. The results 
in Figures 6.57-6.62 suggest that the selected partition coefficient, 
Kqw for atrazine may be too low. The model appears to predict that 
atrazine would leach more rapidly than the observed rate of movement. It 
also appears that atrazine is persistent in the soil and that the 78 day 
half-life used as a model input may also be too low. Figures 6.63-6.68 
clearly show that the half-life selected for alachlor was too low. It is 
difficult to assess the influence of the used for alachlor since the 
low half-life degraded the alachlor so rapidly that very little was 
available to be leached. 




20 40 60 80 100 120 140 160 180 200 

DAYS SINCE 11/12/86 



Figure 6.56. Comparison of percolation volumes predicted by GLEAMS 
and PRZM. 



0.0 0.2 



ATRAZINE CONC. (mg/kg) 



145 



a 



0.4 
-,1 — i L_i 



0.6 0.8 1.0 12 14 



Sample Date: 1 1 /18/86 



MEASURED 





50 \ 

75 i 

100- 

175^ 
200-^ 



Figure 6.57. Carparison of measured and PRZM predicted atrazrn. 
concentrations in the soil on l^aeT 



PRZM 



Notes: 

- No Samples Below 5 cm 

- Measured Value is Ava 

of 14 Samples 



6 




25 

50 i 

75 i 
100 

trj 

E"" 125- 
Q 150 

175 
200 J 



0.0 0.2 

J — i — i — L_i 



ATRAZINE CONC. (mg/kg) 

0.4 or n « 



0-6 0.8 1.0 



Sample Date: 11/24/86 



MEASURED 
PRZM 



Notes: 

- No Samples Below 20 cm 

- Measured Values are Ava ' 

of 8 Samples/Depth 



Figure 6.58. Comparison of measured and FRZM predict ^ • 
concentrations in the soil on7l/^^ 



146 



0.0 
0- 



25 ~ 
50 i 
75 z 
100 

W 

^ 125- 



a 

o 



150-E 
1 75 -EP 
200- 



= 3 

- < 



ATRAZINE CONC. (mg/kg) 

0.1 0.2 0.3 0.4 5 

-I 1 1 1 1 1 1 1 1 1 1 I L_J ' ' ' ■ I ' ■ i i I 

I / Sample Date : 12/22/86 

MEASURED 

PRZM 

/ 

/ 

/ 

Notes: 

— No Samples Above 25 cm. 

— Measured Values are Avg. 

of 6 Samples/Depth 



Figure 6.59. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 12/22/86. 



ATRAZINE CONC. (mg/kg) 





o- 




25^ 

- 




50- 


a 

o 


75- 






100 z _ 


K 




E- 


1 25 - 






DE 


150t 




1 75 ^ 




200- 



0.0 0.1 0.2 0.3 
4 — ' — 1 — 1 — I — i — i — ; — i 1 i i i ' 



0.4 0.5 
-j — i — i — i l i i i i i 



Sample Date: 2/09/87 



MEASURED 
PRZM 



Notes: 

— No Samples Below 135 cm. 

— Measured Values are Avg. 

of 6 Samples/Depth 



Figure 6.60. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 02/09/87. 



ATRAZINE CONC. (mg/kg) 



0.00 




0.05 



253"^ 
50- 



s 



75 ~ 
100- 

« 150^ 



175- 
200- 



0.10 



j i i i_ 



-U 1 I L_ 



-I l_ 



0.15 0.20 
i . I i i i i i 



Sample Date: 3/16/87 

MEASURED 

PRZM 



Notes: 

— Measured Values are Avg. 
of 5 Samples/Depth 



Figure 6.61. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 03/16/87. 



ATRAZINE CONC. (mg/kg) 

0.00 0.05 0.10 0.15 20 

d — 1 — 1 — 1 — I — 1 — 1 — ' — 1 — 1 — 1 — 1 — 1 i i i 



a 

o 
K 

Q 



253 

50 

75 
100 i 
125 
150 
175 3 
200 



Sample Date: 5/25/87 



MEASURED 
PRZM 



Notes: 

— Measured Values are Avg. 
of 7 Samples/Depth 



Figure 6.62. Comparison of measured and PRZM predicted atrazine 
concentrations in the soil on 05/25/87. 



0.0 







B 



w 

E- 1 

a, 



25 
50 
75 
100t 
1 25 t 
150t 
1 75 4 
200- 



0.5 

j L 



ALACHLOR CONC. (mg/kg) 

1.0 15 2,0 2.5 3.0 3.5 



148 



Sample Date: 11/18/86 



MEASURED 
PRZM 



Notes: 

- No Samples Below 5 cm. 

— Measured Value is Avg. 

of 14 Samples 



Figure 6.63. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 11/18/86. 



0.0 

°i 

25 z > 

- i 

50 \! 

- 

75 i 
100- 
125 
150 

,75 j 

200 d 



0.5 



ALACHLOR CONC. (mg/kg) 



1 .0 



1 .5 
. i_ 



2.0 
i 



2.5 



3.0 



3.5 



Sample Date: 11/24/86 



MEASURED 
PRZM 



Notes: 

- No Samples Below 20 cm. 

— Measured Values are Avg. 

of 8 Samples/Depth 



Figure 6.64. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 11/24/86. 



0- 
25-. 
50z, 
75^ 

- 

100^ 
1 25 ^ 
1 50 -E 
175E 
200- 



ALACHLOR CONC. (mg/kg) 



149 



0.0 0.5 1.0 
— i — rJ — i — I — i — i i ' i 



1.5 



j i_ 



2.0 



Sample Date : 12/22/86 



MEASURED 
PRZM 



Notes: 

- No Samples Above 25 cm. 

- Measured Values are Avg. 

of 6 Samples/Depth 

- Alachlor not Detected in 

These Samples 



Figure 6.65. 



Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 12/22/86. 



ALACHLOR CONC. (mg/kg) 



0.0 0.5 1 

-i — i — i — i — i — i i i ■ i i 



25i 
50\ 
75 

100 

1253 

- 

150^ 
175 
200 J 



1.5 2.0 
-j — i — I i i i ■ | 



Sample Date: 2/09/87 



MEASURED 
PRZM 



Notes: 

- No Samples Below 135 cm. 

— Measured Values are Avg. 

of 6 Samples/Depth 



Figure 6.66. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 02/09/87. 



ALACHLOR CONC. (mg/kg) 



150 



0.0 0.5 
4 — i 1 1 i L 



a 

o 

M 
E- 1 



25 p 

50^ 

75 
100^ 
125t 



« 150^ 
175^ 
200- 
Figure 6.67. 



1.0 1.5 
j i i i ' i i i l 



2.0 



i , i ' ' 



Sample Date: 3/16/87 



MEASURED 
PRZM 



Notes: 

— Measured Values are Avg. 
of 5 Samples/Depth 



Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 03/16/87. 



0.0 





25^ 




50 ~ 






a 

o 


75 1 








100^ 


K 




E- 1 


125^ 


cu 




DE 


150^ 




175 7 




200- 



ALACHLOR CONC. (mg/kg) 

0.5 1.0 1.5 2.0 
_i — i — i — I — i i i i l i i i i i i i i i i 



Sample Date: 5/25/87 



MEASURED 
PRZM 



Notes: 

— Measured Values are Avg. 
of 7 Samples/Depth 



Figure 6.68. Comparison of measured and PRZM predicted alachlor 
concentrations in the soil on 05/25/87. 



151 

Figures 6.69-6.71 shew the measured and predicted concentrations of 
bromide in the soil solution at depths of 61, 122, and 183 cm following 
the first application of bromide. Figures 6.72-6.74 show the bromide 
concentrations at the same depths following the second bromide 
application. PRZM predicted concentrations are shown for both free 
drainage and restricted drainage options (see Section 5.2.3 for 
discussion of drainage options in PRZM) . In almost every case shown in 
Figures 6.69-6.74, the time to peak concentration of the measured data 
falls between the predicted time to peak concentration for the two 
drainage options. Thus it would appear that a drainage rate parameter 
could be selected which would match the measured time to peak 
concentration more closely. This would be part of a calibration 
procedure. Calibration of the models was not attempted since there is 
only one year of data from this study and if this data is used for 
calibration, there would be no independent data to test the calibrated 
models against. 

It should also be noted in Figures 6.69-6.74 that in all cases the 
measured and predicted peak concentrations agreed to within an order of 
magnitude, and that in most cases they agreed to within a factor of 2 to 
3. Thus for the case of the bromide applications, PRZM would meet the 
criteria for acceptance suggested by Hedden (1986) . In the work 
presented here PRZM and GLEAMS were essentially run in a screening mode. 
Many of the model parameters were selected from tables and other 
information contained the model's users manuals. Parameters were not 
optimized or calibrated to produce the best fit. At this stage of data 
collection and model use, it was decided to investigate how the models 



152 




DAYS SINCE 11/12/86 



Figure 6.69. Measured and FRZM predicted bromide concentrations in 
the soil solution at a 61 cm depth following the first 
application. 




DAYS SINCE 11/12/86 



Figure 6.70. Measured and PRZM predicted bromide concentrations i 
the soil solution at a 122 cm depth following the 
first application. 



153 



12n 



btf 10- 
8H 



PRZM— FREE DRAINAGE 
PRZM— RESTR. DRAINAGE 
AVG. 183 cm SAMPLERS 




10 



20 30 40 50 

DAYS SINCE 11/12/86 



i i i i i i i 

60 70 



80 



Figure 6.71. Measured and PRZM predicted bromide concentrations in 
the soil solution at a 183 cm depth following the 
first application. 



15-i 




■_■ PRZM— FREE DRAINAGE 

PRZM— RESTR. DRAINAGE 
•— • AVG. 61 cm SAMPLERS 



160 



t — i — f*~\ — i — r 
170 180 190 200 

DAYS SINCE 11/12/86 



210 



Figure 6.72. Measured and PRZM predicted bromide concentrations i 
the soil solution at a 61 cm depth following the 
second application. 



154 



15-, 



PRZM— FREE DRAINAGE 
PRZM-RESTR. DRAINAGE 
AVG. 122 cm SAMPLERS 




DAYS SINCE 11/12/86 



Figure 6.73. Measured and PRZM predicted bromide concentrations in 
the soil solution at a 122 cm depth following the 
second application. 



15-i 



PRZM— FREE DRAINAGE 
•_• PRZM-RESTR. DRAINAGE 
AVG. 183 cm SAMPLERS 




160 



f 'T' T '■' I I ■ I I | I I I I | I | , -r-. 

170 180 190 200 210 

DAYS SINCE 11/12/86 



Figure 6.74. Measured and PRZM predicted bromide concentrations in 
the soil solution at a 183 cm depth following the 
second application. 



155 

would perform when run with minimal site-specific inputs, a 
governmental regulator would certainly not have detailed site specific 
data available for all the combinations of crop/soil/chemical that may be 
of interest. Such a user would be unlikely to have site measured 
application rates for use as input. If the normal application rates for 
atrazine and alachlor are 4.9 kg/ha on the cropping system which is being 
simulated, the model user would certainly use those values as the input 
to the model. They would not know that the actual average application 
rates may only be 3.4 kg/ha. 

A small number of soil solution samples from samplers were extracted 
for analysis of atrazine residues. The volume of these samples was 
between 30 and 40 mL. Figure 6.35 shows the measured and PRZM simulated 
concentrations of atrazine in the soil solution at a depth of 61 cm. PRZM 
was run in the free drainage mode using the calculated plant uptake 
coefficient of 0.65. 



0.6 h 



SOIL SOLUTION SAMPLES 
61 cm DEPTH 



09N-2 



_ 10S-2 
10N-2 
PRZM — FD 




n o.i q 
■< 

OS 

r , 

^j* 0.0 i i i — i — | — i — i — i — i — i — i — i — i — i — i — i — i — i — r 

20 40 60 80 

DAYS SINCE 11/12/86 

Figure 6.75. Measured and PRZM predicted atrazine concentrations in 
the soil-water at a depth of 61 cm. 



i I i — i — i — j — t 
100 120 



156 

The maximum observed c»ncentration of atrazine at the 61 can depth 
was approximately 0.35 mg/L and the maximum concentration predicted by 
PRZM was 0.57 mg/L. The predicted and observed maximum concentrations 
are again within a factor of 2. The 0.35 mg/1 maximum observed value 
however was from one of three samplers from which seme samples were 
extracted. The data from the other two samplers do not shew any distinct 
peak concentration, however, the concentrations are within an order of 
magnitude of the predicted values. The models were run with the intended 
application rate of 4.9 kg/ha, but the actual application samples suggest 
the true application rate was 3.4 kg/ha. The time to peak concentration 
between predicted and observed was reasonably close; however PRZM 
simulated a much broader peak than was observed. 

The simulated transport of applied chemicals by the two models were 
in reasonable agreement with each other. The differences are probably 
due to the detail of discretization of the root zone and the method of 
calculated chemical transport as discussed previously. PRZM. can more 
nearly match the observed soil layering than can GLEAMS due to the fact 
that GLEAMS describes the entire root zone with only seven layers. PRZM 
and GLEAMS both simulated large losses of atrazine from the root zone (40 
and 62% of application, respectively) . The PRZM simulated transport of 
atrazine to a depth of 2.6 m amounted to 24% of the mass applied. As 
discussed previously, less than 4% of the applied atrazine could be 
accounted for on a given day within the saturated zone. This figure 
does not account for storage in the vadose zone or mass transported off 
site. 



CHAPTER 7 
SUMMARY AND CONCLUSIONS 



A field experiment was conducted to observe the movement of two 
surface applied herbicides (atrazine and alachlor) through the soil 
profile and within a shallow water table aquifer. Bromide was applied to 
the soil surface to act as a non-adsorbed tracer of water movement. The 
nitrate component of a surface applied fertilizer was also monitored 
through the soil profile and within the water table. 

Instrumentation was installed on a 0.7 ha field for collection of 
soil water samples from the unsaturated zone and water samples from the 
saturated zone. Soil water samples were collected with soil solution 
samplers installed in six groups of three samplers each at depths of 61, 
122, and 183 cm. Shallow groundwater samples were collected from 5 cm 
diameter PVC monitoring wells. The wells were installed on a 12 by 12 m 
grid over the study site. Soil samples were collected several times 
during the study from various depths beneath the application area. 

Methods were developed to extract the herbicide residues from soil 
and water samples. Herbicide residue samples were analyzed on a gas 
chromatograph using a nitrogen-phosphorus detector. Concentrations of 
the inorganic tracers were analyzed using an ion chromatograph. 

The uniformities of application of atrazine, alachlor, and bromide 
were measured. There was considerable variability observed in the 
application rates (kg/ha) of the three chemicals. The chemicals were 

157 



158 

applied using a boon sprayer developed for use in chemigation research. 
The results indicate that chemical application rates from conventional 
farm equipment are likely to be highly variable. The variability of 
application, in addition to the variability of soil properties, makes 
prediction of field measured concentrations difficult. 

Concentrations of bromide and nitrate at all three monitored depths 
showed a high degree of variability. The peak concentration and the time 
required to reach the peak concentration varied significantly between 
sampling locations. The effect of this variability may be to reduce 
maximum concentrations and increase the duration of loading reaching a 
water table. 

Atrazine moved rapidly through the sandy soil on the study site. 
Concentrations of atrazine in the soil water at a depth of 61 cm reached 
0.35 mg/L 19 days after application. Detectable levels of atrazine 
reached the water table 2 months after application. Atrazine 
concentrations as high as 0.09 mg/L were observed in the groundwater 
nearly 6 months after application. 

Alachlor was not detected in the soil below a depth of 45 cm. No 
trace of alachlor was detected in the groundwater samples. 

As much as 20 percent of the nitrate from the fertilization of the 
study site was observed to be in the groundwater on a given date. The 
decreasing concentrations of nitrate in the groundwater with continued 
percolation of water suggests that most of the nitrate was leached from 
the soil profile. 

The PRZM. and GLEAMS models were found to be easy to use. Sufficient 
information is provided in the user manuals for estimation of required 



159 

parameters. When the models were run using pesticide properties obtained 
from the manuals, simulated leaching and degradation of the herbicides 
exceeded field observations. The models were not calibrated to the 
observed data. 

The data collected during this study provide a picture of the 
concentrations of various chemicals within the soil profile and 
groundwater on specific dates. Data to accurately quantify the movement 
of water on the site, however, are missing. The models predict the mass 
flux of chemicals past a given point (usually the bottom of the root 
zone) . The observed data from this first year study are insufficient to 
determine mass fluxes and therefore can not be used to adequately test 
the mass flux predictions of the models. PRZM simulation results were 
used to compare observed and predicted concentrations of the chemicals. 
This is because PRZM will report the simulated concentrations at any 
depth on a daily basis. GLEAMS will only output soil concentration data 
(Mg/g) on days with a storm event that causes leaching below the root 
zone. This makes it difficult to compare GLEAMS predictions with samples 
taken between storm events. 

The results presented in this dissertation represent data from the 
first year of a field study and application of two pesticide transport 
models to the conditions present at the field site during the study. 
Much has been learned about sample collection, sample analysis, and the 
limitations of the data which have been collected, for both describing 
the movement of the chemicals and testing model predictions. 



CHAPTER 8 

REXXMMENDATICNS FOR IMPROVEMENTS AND FURTHER STUDY 



Additional information must be collected in order to quantify the 
mass flux of chemicals and water within the vadose zone and the 
groundwater. 

The water content, or tension, of the soil must be observed on a 
more frequent basis. Soil moisture blocks or tensiometers with pressure 
transducers could be read on frequent intervals (5 min - 1 hr) using a 
datalogger. If there is sufficient sensitivity to the small changes in 
water content on this soil (0.13 - 0.02 cm 3 /cm 3 ) then the total flux of 
water moving by the tensiometers could be estimated. 

Further characterization of the restricting layer is needed. Core 
samples should be taken from throughout the soil profile and presumed 
restricting layer. Measurements of hydraulic conductivity, bulk density, 
particle size distribution, and organic matter content should be made on 
the cores. These data would refine estimates of soil properties 
throughout the soil profile and further define the location of the 
restricting layer. A detailed mapping of the surface of the restricting 
layer using ground penetrating radar (GPR) would help to define the 
locations and extent of small channels or irregularities. 

Knowing where the bottoms of the observation wells are located 
relative to the restricting layer will improve the estimates of flow 
depth. Improved values for conductivity and the cross-sectional area of 

160 



161 

flew at a location will improve calculations of water flux. An estimate 
of the seepage of water through the restricting layer may also improve 
the water budget for this site. The current system adequately defines 
the elevation of the water table surface from which gradients are 
calculated. 

Model validation will require (in addition to the soil properties 
listed above) that the partition coefficient and degradation rate for 
the chemical of interest be measured from on-site soil samples. The 
accuracy of model predictions of mass fluxes should improve significantly 
by using site-specific data. Careful measurement of application rates 
should be made. Measured application rates were lower than the intended 
rates. The models should also be tested using the observed range of 
sensitive soil and chemical parameters. 

A second year study on this same site should be done to try to 
quantify some of the characteristics or properties described above. In 
addition, the study should monitor the movement of atrazine, one other 
pesticide, bromide, and nitrate. The usefulness of the second year data 
would reflect and be enhanced by the lessons learned from the work done 
to complete this dissertation. 



REFERENCES 

Agrichemical Handbook, The. 1983. Royal Society of Chemistry. The 
University, Nottingham, England. 

Alexander, W J. and S. K. Little. 1986. Ground water vulnerability 
assessment in support of the first stage of the national pesticide 
survey, in Agricultural Impacts On Groundwater Quality — A 
conference. Omaha, NE. National Water Well Association, Water Well 
Publishing Co., Dublin, OH. pp. 77-86. 

Aller, L. , T. Bennett, J. H. Lehr, and R. J. Petty. 1985. DRASTIC: A 
standardized system for evaluating ground water pollution 
potential using hydrogeologic settings. U. S. Environmental 
Protection Agency. Washington, DC. EPA 600/2-85/018. 

Asmussen, L. E., H. F. Perkins, and H. D. Allison. 1986. Subsurface 
descriptions by ground penetrating radar for watershed 
delineation. Georgia Agricultural Experiment Stations Research 
Bulletin No. 340. Athens, GA. 15 pp. 

Berteau, P. E. and D. P. Spath. 1986. The toxicolcgical and 

epidemiological effects of pesticide contamination in California 
ground water, in Evaluation of Pesticides in Groundwater. W. Y. 
Garner, R. C. Honeycut, and H. N. Nigg, Eds. ACS Symposium Series 
No. 315, American Chemical Society, Washington D.C. pp. 423-435. 

Brinsfield, R. B. , K. W. Staver, and W. L. Magette. 1987. Impact of 
tillage practices on pesticide leaching in Coastal Plain soils. 
ASAE Paper No. 87-2631. ASAE, St. Joseph, MI. 21 pp. 

Brinsfield, R. B. , K. W. Staver, and W. L. Magette. 1988. The role 
of cover crops in reducing nitrate leaching to groundwater, in 
Proceedings of the Agricultural Impacts On Groundwater Quality — A 
conference. Des Moines, IA. National Water Well Association. Water 
Well Publishing Co., Dublin, OH. pp. 127-146. 

Carlisle, V. W. , R. E. Caldwell, F. Sodek, III., L. C. Hammond, F. 
G. Calhoun, M. A. Granger, and H. L. Breland. 1978. 
Characterization data for selected Florida soils. University of 
Florida, Institute of Food and Agricultural Sciences, Soil 
Science Department Research Report No. 78-1. Gainesville, FL. 
335 pp. 



162 



Carsel, R. F. , C. N. Smith, L. A. Mulkey, J. D. Dean, and P. P. 
Jowise. 1984. Users manual for the pesticide root zone model 
(PRZM) release 1. U. S. Environmental Protection Agency, Athens, 
GA. EPA-600/3-84-109. 

Carson, R. L. 1962. Silent Spring. Houghton Mifflin, Boston, MA. 

Cheng, H. H. and W. C. Roskinen. 1986. Processes and factors 
affecting transport of pesticides to ground water, in Evaluation 
of Pesticides in Groundwater. W. Y. Garner, R. C. Honey cut, and H. 
N. Nigg, Eds. ACS Symposium Series No. 315, American Chemical 
Society, Washington D.C. pp. 170-196. 

Cohen, S. Z., C. Eiden, and M. N. Lorber. 1986. Monitoring ground 
water for pesticides, in Evaluation of Pesticides in Groundwater. 
W. Y. Garner, R. C. Honeycut, and H. N. Nigg, Eds. ACS Symposium 
Series No. 315, American Chemical Society, Washington D.C. 
pp. 170-196. 

Cooper, S. C. 1986. Design and installation of a monitoring 
network for measuring the movement of aldicarb and its residues 
in the unsaturated and saturated zones, Lee County, Georgia, in 
Agricultural Impacts On Groundwater Quality — A conference. 
Omaha, NE. National Water Well Association. Water Well Publishing 
Co., Dublin, OH. pp. 194-223. 

Dean, J. D. , P. P. Jowise, and A. S. Donigian, Jr. 1984. Leaching 
Evaluation of Agricultural Chemicals (LEACH) Handbook. U. S. 
Environmental Protection Agency, Athens, Georgia. 
EPA-600/3-84-068. 407 pp. 

Dean, J. D. , and R. F. Carsel. 1988. A linked modeling system for 
evaluating impacts of agricultural chemical use. in Agricultural 
Impacts On Groundwater Quality — A conference. Des Moines, IA. 
National Water Well Association. Water Well Publishing Co. , 
Dublin, OH. pp. 477-502. 

Dick, W. A., W. M. Edwards, and F. Haghiri. 1986. Water movement 
through soil to which no-tillage cropping practices have been 
continuously applied, in Agricultural Impacts On Groundwater 
Quality — A conference. Omaha, NE. National Water Well 
Association. Water Well Publishing Co., Dublin, OH. pp. 243-252. 

Donigian, A. S., Jr. and P. S. C. Rao. 1986a. Overview of 

terrestrial processes and modeling, in Vadose Zone Modeling of 
Organic Pollutants. S. C. Hern and S. M. Melancon, Eds. Lewis 
Publishers, Inc. Chelsea, Ml. pp. 3-35. 



164 



Donigian, A. S., Jr. and P. S. C. Rao. 1986b. Example model testing 
studies, in Vadose Zone Modeling of Organic Pollutants. S. C. 
Hern and S. M. Melancon, Eds. lewis Publishers, Inc. Chelsea, MI. 
pp. 103-131. 

Enfield, C. G. , R. F. Carsel, S. Z. Cohen, T. Phan, and D. M. 
Walters. 1982. Approximating pollutant transport to ground water. 
Ground Water. 20:711-722. 

Farm Chemical Handbook. 1984. Meister Publishing Company, 
Willoughby, CH. 

Freeze, R. A. and J. A. Cherry. 1979. Groundwater. Prentice-Hall, 

Englewood Cliffs, NJ. 604 pp. 

Garner W. Y. , R. C. Honeycut, and H. N. Nigg, Eds. 1986. Evaluation of 
Pesticides in Groundwater. ACS Symposium Series No. 315, American 
Chemical Society, Washington D.C. pp. 100-115. 

Gish, T. J. 1987. Bromide transport in structured soils. ASAE Paper 
No. 87-2625. ASAE, St. Joseph, MI. 10 pp. 

Gish, T. J., C. S. Helling and P. C. Kearney. 1986. Simultaneous 
leaching of bromide and atrazine under field conditions, in 
Agricultural Impacts On Groundwater Quality — A conference. 
Omaha, NE. National Water Well Association. Water Well Publishing 
Co., Dublin, CH. pp. 286-297. 

Golden Software, Inc. 1987. Surfer Reference Manual. Golden, CO. 
329 pp. 

Hallberg, G. R. , 1986. Overview of agricultural chemicals in ground 
water, in Agricultural Impacts On Groundwater Quality — A 
conference. Omaha, NE. National Water Well Association. Water 
Well Publishing Co., Dublin, CH. pp. 1-63. 

Hedden, K. F. 1986. Example field testing of soil fate and 

transport model, PRZM, Dougherty Plain, Georgia, in Vadose Zone 
Modeling of Organic Pollutants. S. C. Hern and S. M. Melancon, 
Eds. lewis Publishers, Inc. Chelsea, MI. pp. 81-101. 

Helling, C. S. and T. J. Gish. 1986. Soil characteristics affecting 
pesticide movement into groundwater, in Evaluation of Pesticides 
in Groundwater. W. Y. Garner, R. C. Honeycut, and H. N. Nigg, 
Eds. ACS Symposium Series No. 315, American Chemical Society, 
Washington D.C. pp. 14-38. 

Hetrick, D. M. and C. C. Travis. 1988. Model predictions of 
watershed erosion components. Water Resources Bulletin. 
24(2)413-419. 



165 



Hook, J. E. 1985. Irrigated corn management for the coastal plain. 
Georgia Agricultural Experiment Stations Research Bulletin 
No. 335. Athens, GA. 34 pp. 

Hook, J. E. 1987. Soil physical properties of a pesticide study 
site in the Dougherty Plain of Georgia. Final Report. U. S. 
Environmental Protection Agency, Athens, GA. 

Hornsby, A. G. , P. S. C. Rao, W. B. Wheeler, P. Nkedi-Kizza, and R. 
L. Jones. 1983. Fate of aldicarb in Florida citrus soils: 1. 
field and laboratory studies, in Proceedings of the 
Characterization and Monitoring of the Vadose (Unsaturated) Zone. 
Las Vegas, NV. D. M. Nielson and M. Carl, Eds. National Water 
Well Association. Water Well Publishing Co. , Dublin, OH. 
pp. 936-958. 

Jones, R. L. , A. G. Hornsby, P. S. C. Rao, and M. P. Anderson. 
1987. Movement and degradation of aldicarb residues in the 
saturated zone under citrus groves on the Florida Ridge. J. 
Contaminant Hydrology. 1:265-285. 

Jones, R. L. , P. S. C. Rao, and A. G. Hornsby. 1983. Fate of 
aldicarb in Florida citrus soils: 2. model evaluation, in 
Proceedings of the Characterization and Monitoring of the Vadose 
(Unsaturated) Zone. Las Vegas, NV. D. M. Nielson and M. Carl 
Eds. National Water Well Association. Water Well Publishing 
Co., Dublin, OH. pp. 959-978. 

Jury, W. A. 1982. Simulation of solute transport using a transfer 
function model. Water Resources Research. 18(2) : 363-368. 

Jury, W. A. 1985. Spatial Variability of Soil Physical Parameters in 
Solute Migration: A critical Literature Review. Electric Power Research 
Institute. Palo Alto, CA. EPRI EA-4228. 

Jury, W. A. 1986a. Mathematical derivation of chemical transport 
equations, in Vadose Zone Modeling of Organic Pollutants. S. C. 
Hern, and S. M. Melancon, Eds. Lewis Publishers, Inc. Chelsea, 
MI. pp. 271-288. 

Jury, W. A. 1986b. Chemical movement through soil, in Vadose Zone 
Modeling of Organic Pollutants. S. C. Hern, and S. M. Melancon, 
Eds. Lewis Publishers, Inc. Chelsea, MI. pp. 135-158. 

Jury, W. A. 1986c. Adsorption of organic chemicals onto soil, in 
Vadose Zone Modeling of Organic Pollutants. S. C. Hern, and S. M. 
Melancon, Eds. Lewis Publishers, Inc. Chelsea, ME. pp. 177-189. 

Jury, W. A. 1986d. Volatilization from soil, in Vadose Zone 
Modeling of Organic Pollutants. S. C. Hern, and S. M. Melancon, 
Eds. Lewis Publishers, Inc. Chelsea, MI. pp. 159-176. 



166 



Jury, W. A. , H. Elabd, and M. Resketo. 1986. Field study of 
napropamide movement through unsaturated soil. Water Resources 
Research. 22(5)749-755. 

Jury, W. A. and R. L. Valentine. 1986. Transport mechanisms and 
loss pathways for chemicals in soils, in Vadose Zone Modeling of 
Organic Pollutants. S. C. Hern, and S. M. Melancon, Eds. Lewis 
Publishers, Inc. Chelsea, ME. pp. 37-60. 

Kelley, R. , G. R. Hallberg, L. G. Johnson, R. D. Libra, C. A. 
Thompson, R. C. Splinter, and M. G. DeTroy. 1986. Pesticides in 
ground water in Iowa, in Agricultural Impacts On Groundwater 
Quality — A conference. Omaha, NE. National Water Well 
Association. Water Well Publishing Co., Dublin, OH. pp. 662-647. 

Khisel, W. G. (Ed) . 1980. CREAMS: A Field-Scale Model for 
Chemicals, Runoff, and Erosion from Agricultural Management 
Systems. U.S. Department of Agriculture, Conservation Research 
Report No. 26, Washington, DC. 640 pp. 

Leonard, R. A. 1988. Herbicides in surface waters, in Environmental 
Chemistry of Herbicides, Volume 1. R. Graver, Ed. CRC Press, Boca 
Raton, FL. pp. 45-87. 

Leonard, R. A. and W. G. Khisel. 1987. Personal Communication. Soil 
Scientist and Hydraulic Engineer, respectively. USDA, Southeast 
Watershed Research Laboratory, Tifton, GA. 

Leonard, R. A. and W. G. Khisel. 1988. Can pesticide transport models be 
validated using field data: now and in the future? J. Env. Sci. and 
Health, Part B: Pesticides, Food Contaminants and Agricultural Wastes. 
In Press. 

Leonard, R. A., W. G. Khisel, and D. A. Still. 1987. GLEAMS: 
Groundwater loading effects of agricultural management systems. 
Transactions of the ASAE. 30(5) : 1403-1418. 

Libra, R. D. , G. R. Hallberg, and B. E. Hoyer. 1987. Impacts of 
agricultural chemicals on ground water quality in Iowa, in Ground 
Water Quality and Agricultural Practices. D. M. Fairchild, Ed. 
Lewis Publishers, Chelsea, ME. pp. 185-216 

Libra, R. D., G. R. Hallberg, B. E. Hoyer, and L. G. Johnson. 
1986. Agricultural impacts on groundwater quality: the Big 
Spring Basin study, in Agricultural Impacts On Groundwater 
Quality—A conference. Omaha, NE. National Water Well 
Association. Water Well Publishing Co., Dublin, OH. pp. 253-273. 

Litaor, M. I. 1988. Review of soil solution samplers. Water 
Resources Research. 24(5) : 727-733. 

Marti, L. R. 1987. Personal communication. Chemist, USDA, Southeast 
Watershed Research Laboratory, Tifton, GA. 



167 



Marti, L. R. , J. de Kanel, and R. C. Dougherty. 1984. Screening 
for organic contamination of groundwater: ethylene dihromide in 
Georgia irrigation wells. Environ. Sci. Technol. 18(12) : 973-974. 

National Water Well Association. 1986. Proceedings of the 
Agricultural Impacts On Groundwater Quality— A conference. 
Omaha, NE. Water Well Publishing Co. , Dublin, OH. 685 pp. 

National Water Well Association. 1988. Proceedings of the 
Agricultural Impacts On Groundwater Quality— A conference. Des 
Moines, IA. Water Well Publishing Co. , Dublin, OH. 658 pp. 

Pionke, H. B., D. E. Glotfelty, A. D. Lucas, and J. B. Urban. 1988. 
Pesticide contamination of groundwaters in the Mahantango Creek 
Watershed. J. Environmental Quality. 17(1): 76-84. 

Rao, P. S. C. and J. M. Davidson, (Eds.). 1980. Retention and 
transformations of selected pesticides and phosphorus in 
soil-water systems: a critical review. U. S. Environmental 
Protection Agency, Athens, GA. EPA-600/3-82-060. 

Rao, P. S. C. , D. E. Rolston, R. E. Jessup, and J. H. Davidson. 
1980. Solute transport in aggregated porous media: theoretical 
and experimental evaluation. Soil Sci. Sec. Amer. J. 
44(6) : 1139-1146. 

Rao, P. S. C. , A. G. Hornsby, and R. E. Jessup. 1985. Indices for 
ranking the potential for pesticide contamination of groundwater. 
Soil and Crop Sci. Sec. Florida Proc. 44:1-8. 

Rao, P. S. C. , K. S. V. Edvardsson, L. T. Ou, R. E. Jessup, 

P. Nkedi-Kizza, and A. G. Hornsby. 1986. Spatial variability of 
pesticide sorption and degradation parameters, in Evaluation of 
Pesticides in Groundwater. W. Y. Garner, R. C. Honeycut, and H. 
N. Nigg, Eds. ACS Symposium Series No. 315, American Chemical 
Society, Washington D.C. pp. 100-115. 

Rao, P. S. C, R. E. Jessup, and J. M. Davidson. 1988. Mass flow 
and dispersion, in Environmental Chemistry of Herbicides, Volume 
1. R. Graver, Ed. CRC Press, Boca Raton, FL. pp. 21-44. 

Ritter, W. F. 1986. Pesticide contamination of groundwater — a 
review. ASAE Paper No. 86-2028. ASAE, St. Joseph, MI 49085. 

Rohde, W. A. , L. E. Asmussen, E. W. Hauser, M. L. Hester, and H. D. 
Allison. 1981. Atrazine persistence in soil and ■transport in 
surface and subsurface runoff from plots in the coastal plain of 
the southern United States. Agro-Ecosystems. 7(1981) : 225-238. 



168 



Shoemaker, L. L. and W. L. Magette. 1987. Ground water models for 
assessing agricultural best management practice. ASAE Paper 
No. 87-2021. ASAE, St. Joseph, MI 49085. 

Skaggs, R. W. 1978. A water management model for shallow water 
table soils. Water Resources Research Institute of the University 
of North Carolina. Report No. 134. 178 pp. 

Smith, C. N., W. R. Payne, Jr., L. A. Mulkey, J. E. Benner, R. S. 
Parrish, and M. C. Smith. 1981. The persistence and disappearance 
by washoff and dryfall of methoxychlor from soybean foliage — a 
preliminary study. J. Env. Sci. and Health. B16 (6) : 779-794. 

Smith, C. N. and R. F. Carsel. 1986. A stainless steel soil 
solution sampler for monitoring pesticides in the vadose zone. 
Soil Sci. Sec. Am. J. 50(1) : 263-265. 

Smith, R. E. and J. R. Williams. 1980. Simulation of the surface water 
hydrology, in Knisel, W. G. (Ed) . CREAMS: A Field-Scale Model for 
Chemicals, Runoff, and Erosion from Agricultural Management 
Systems. U.S. Department of Agriculture, Conservation Research 
Report No. 26, Washington, DC. 640 pp. 

Stansell, J. R. , C. L. Butts, K. A. Harrison, and J. C. Garner. 
1982. Irrigation system efficiency survey for Georgia. 
Environmental Resources Center, Georgia Institute of Technology, 
Atlanta, GA. ERC 10-82. 32 pp. 

Steenhuis, T. S., S. Pacenka, and K. S. Porter. 1987. MOUSE: a 
management model for evaluating groundwater contamination from 
diffuse surface sources aided by computer graphics. Applied Agr. 
Research. 2(4) : 277-289. 

Sumner, H. R. , E. D. Threadgill, J. R. Young, and D. L. Cochran. 
1987. Irrigation system simulator for small plot research. ASAE 
Paper No. SER-87-201. ASAE, St. Joseph, MI 49085. 

U. S. Department of Agriculture, Soil Conservation Service. 1972. 
SCS National Engineering Handbook, Section 4, Hydrology. 
Washington, DC. 458 pp. 

U. S. Environmental Protection Agency. 1987a. Agricultural 

Chemicals in Ground Water: Proposed Pesticide Strategy. Office of 
Pesticides and Toxic Substances, Washington, DC. 150 pp. 

U. S. Environmental Protection Agency. 1987b. Environmental News. 
Office of Public Affairs, Washington, DC. Feb. 4, 1987. 

Valentine, R. L. 1986. Nonbiological transformation, in Vadose Zone 
Modeling of Organic Pollutants. S. C. Hern, and S. M. Melancon, 
Eds. lewis Publishers, Inc. Chelsea, MI. pp. 223-244. 



169 



Valentine, R. L. and J. L. Schnoor. 1986. Biotransformation, in 
Vadose Zone Modeling of Organic Pollutants. S. C. Hern, and S. 
M. Melancon, Eds. lewis Publishers, Inc. Chelsea, MI. 
pp. 191-222. 

Voznakova, M. P. and V. Tatar. 1983. Determination of triazines in 
water by GC and LC. J. Chromatographic Science. 21:39-42. 

Wagenet, R. J. 1986. Principles of modeling pesticide movement in 
the unsaturated zone, in Evaluation of Pesticides in Groundwater. 
W. Y. Garner, R. C. Honeycut, and H. N. Nigg, Eds. ACS Symposium 
Series No. 315, American Chemical Society, Washington D.C. 
pp. 170-196. 

Wagenet, R. J., and J. L. Hutson. 1986. Predicting the fate of 
nonvolatile pesticides in the unsaturated zone. J. Environ. Oual. 
15:315-322. 

Wang, H. F. , and M. P. Anderson. 1982. Introduction to Groundwater 
Modeling, Finite Difference and Finite Element Methods. W. H. 
Freeman and Company, San Francisco, CA. 233 pp. 

Wartenberg, D. 1988. Groundwater contamination by Temik aldicarb 
pesticide: the first 8 months. Water Resources Research. 
24(2) : 185-194. 

Wauchope, R. D. 1978. The pesticide content of surface water 
draining from agricultural fields - a review. J. Environ. Qual. 
7(4) : 459-472. 

Wheeler, W. B. 1987. Personal communication. Professor of Food Science 
and Human Nutrition. University of Florida, Gainesville, FL. 

Wischmeier, W. H. , and D. D. Smith. 1978. Predicting rainfall 
erosion losses. U. S. Department of Agriculture, Agriculture 
Handbook No. 537. Washington, DC. 58 pp. 



APPENDIX A 
MONITORING WELL STATISTICS 



171 



CO 

HOHCSCOVOCOTOVOnO\Ht^CV)COOgr^HHr^COlf)rHCOOOr^OV£) 
hOHOOlri(MO^MJ\HrH£)Oa)^Tf(Nfl\a)rlOnrHOOO 

r^wcocar^t^t^r^r^vor^cocot^r^vovor^voininvot^inincoinin 

CMCVJCMCNIOJCNJCNIOJCNICNICMCMCNICNICMCNICNICMCMCNCNICSCM 



8 

8 



cMO^HovoincMco^cocvicoco^^inr^cocricNir^cri 

<*fNfnNNNrlOOOHOh\OIOin«OOOH 



H VO CM CO 

in t}< 



oooooooooooooooio^oxrimmoimijieoooooioeo 

in^onvovovoin^vocMVD'tcjiaivOCTic^r^r^^cocorHiocNjcovD 
coco>o^cor^r^r^voinininvovocMHoot^inininvocMO<rio^co 

OOOOOOOOOOOOOOOOOO0\010l(MMJ>ffiC0(tlt0 

^''fWHeoorii^conninr^i^vo^'Ncoioovocot^co^inno 

HTOinCTvin^COHHOOlO'S'^'OCNlCOVOVOirilOVOCOVOCOVOCNlCTlCO 

niNCfNnriMnnnNnNnnrinojnnpiricMnnoriri 

H 



ininiTi't^cotofflnnnnnoitNNNNHHriHHooooo 
HHHnininininr^r^c^r^>cr><TicT>c^cr\HHHHrHcococococo 

nnnint~-oooai(jicnoicnHHriHH^<t^'j^\o^you)io 

t^r^t^CO<TlOOOOOOOOCNCMCMCMCMCOCOCOCOCO'3«'*'<i«'*J<Tj« 
HHHHHHHi--IHHt--IHr-li--li--IHHHt--li--li--lr-IH 



oco^cocrioocot^^cocor^^oi^rocor^CTicooor^vocrioooocor^ 
onmmHnirihHninhOiOHriinhHninhffirin^inh 

O'tovocg^tficoN^^owooog'tiocoN't^cooiN'i'd^eo 



inuiinr^uivovovovoinvovoincoinvovovovoinvoninin*£vovovo 
cooocooococococacocococococococncocococococococococoaDco 

r>r>r>c>r>^>c>c>r>c>^r»r»>»r»r>r>r»^» 
co in m in in mm co in in in in in in h in in 



SB 



oooooooo 
III I I I IT 



I I I I 

SSdS^9929t992J OHVO ^ c ocrio>cocriOHr^ooQcrio 

OOrHOOOOi— ^0(~io-j-J(--ii— ><-><— i_i<— i i 3*C Jfcl z5 



H 
I 



I I 



. H H 
I I I 



I I 



I I 



OOHOOOOHOOO 



OOOOi-tOOOrHHOO OH 



172 



S S3 



I — I 

S3 



8^ 



X 



1 



H 







Cn 


CO 


vo 


in 


rH 


rH 




co 




o> 




P* 
























CO 


CN 


d 


CN 


CN 


CN 


CN 


CN 


CN 


CN 


CTi 


IT) 


CO 


in 


CN 


in 


rH 


O 


CO 


VO 


in 




co 


H 


• 


• 


• 


• 


• 


• 


• 


CO 


r* 






r- 


VO 


CN 


CN 


CN 


CN 


CN 


CN 


CN 


CN 


CN 


UJ 






*T 


rH 


co 


VO 


1*1 


H 




H 


CO 


CO 


• 


• 


• 


• 


• 


• 


• 


CO 


CO 


CO 




co 




CN 




CN 


CN 


CN 


CN 


CN 


CN 


in 


rH 


CN 




co 


VO 


p* 






CN 


Li ) 


rH 


co 


vo 






CO 


CO 






rH 


o 


o 


o 


o 


cn 


CO 


CO 


LD 


IT) 


IT) 




vo 


CO 


rH 






• 


• 


• 


• 


• 


CO 


CO 


CO 


CO 


o 


CN 


o 


IT) 


ID 


in 


in 






o 


H 




rH 


H 


rH 


H 


CN 


cn 


CO 


co 




o 


in 


CN 




co 


in 




o 


rH 


vo 


• 


• 


• 


• 


• 


• 


• 


CM 




vo 


co 


o 


in 


cn 


rH 


CN 


CO 






CN 


n 



vo vo vo in in in in 
co co co co co co co 

c^c^i^r^ 

CN CN CN \ W\ 
\\\M CN CN CN 

in in in 



$$$$ 

BBBB 



^ ^ ^ y 



'S' m 1 in in 

H H H H H H 

I I I I I I 

r» co cn o vo co 

O O O rH O O 



rH CN CO 



APPENDIX B 
HOW TO GET COMPLETE DATA SET 



Complete data sets showing the concentrations of all chemicals in 
both water and soil samples can be obtained by writing to: 

Matt C. Smith 

Agricultural Engineering Department 

University of Georgia 

P.O. Box 748 

Tifton, Georgia 31793 

The data sets can be made available on tape, or floppy disk (5-1/4 
or 3-1/2 inch) . Data sets will be in ASCII format. 

An example of the water sample data set is given in Table B.l The 
water sample data set also contains data on the elevation of the water 
table at each well that was sampled. This data set contains over 900 
records. An example of the soil sample data set is presented in Table 
B.2. This data set contains approximately 240 records. 

Table B.l Example listing of the water sample data set. 

Sample Station Water Table Chemical Concentration (mg/1) 

Date ID Elevation 

(m) Atrazine Bromide Nitrate Chloride 



5/25/87 


08- 


-09 


27 


.46 


0.006 


0.01 


0.47 


13 


3 


5/25/87 


08- 


-10 


27 


.26 


0.005 


0.04 


3.52 


41 


5 


5/25/87 


08- 


-11 


26 


.92 


0.013 


0.00 


0.00 


82 


3 


5/25/87 


08- 


-12 


25 


.97 


1 


• 


• 






5/25/87 


08- 


-13 


25 


.60 


0.011 


0.02 


0.11 


22. 


87 


•■•No sample 



174 



Table B.2 Example listing of the soil sample data set. 



Sample 
Date 


Sample 
location 


Depth 
(cm) 


Concentration 
Atrazine 


(mg/kg) 
Alachlor 


5/25/87 


1 


0-6 


0.04 


0.02 


5/25/87 


2 


0-6 


0.05 


0.12 


5/25/87 


3 


0-6 


0.03 


0.27 


5/25/87 


4 


0-6 


0.04 


0.17 


5/25/87 


5 


0-6 


0.04 


0.06 


5/25/87 


6 


0-6 


0.04 


0.11 


5/25/87 


7 


0-6 


0.04 


0.12 



APPENDIX C 
WATER BALANCE PROGRAMS 



Figure C.l Listing of program to calculate water and chemical fluxes and 
storages. 



CCCCCCC 

C PROGRAM TO CALCULATE WATER AND CHEMICAL FLUXES AND STORAGES FROM 

C GRIDDED WATER TABLE ELEVATION DATA 

C WRITTEN: 11/22/87 BY: MATT C. SMITH 

C LAST UPDATE: 4/12/88 

CCCCCC 

FROGRAM ANALYZE 

REAL WT(23,45) ,WTMIN,WTMAX,MEANWTD(7) ,KD,BULKDEN 

REAL MAGNITUDE (23, 45) ,DELX,DELY,ANGLE(23,45) ,MAGMEN,MAGMAX 

REAL SLPX(23,45) ,SLPY(23,45) , DIRECTION (2 3, 4 5) ,SURFMAX 

REAL SURF(23,45) ,SURFMIN,ROW,COL,MAXST0R(23,45) ,DRVOL(23,46) 

REAL TMP(23,45) ,XMIN,XMAX,YMIN,YMAX,IMFMAX,IMPMm,XPOS,YFOS 

REAL WTD(23,45) ,WITKIN,WTDMAX,SII : XMAX,SI^MAX,^ 

REAL VELX(23,45) ,VELY(23,45) / FLOWXMAX, FDOWYMAX,FICWXMIN,FIXWYMIN 

REAL FLOWX(23,45) ,FLOWY(23,45) ,BFLOWXl(7) ,BFLOWX2(7) ,BFLOWY(7) 

REAL NETFLOW(7) ,AREA(7) ,COND(45) ,P0RO,NODECOND(7) 

REAL WMASS(23,45) ,CMASS(23,45) ,CONC(23,45) ,TCMASS(7) ,CFLOWYMIN 

REAL BCFLOWXl(7) ,BCFLOWX2(7) ,BCFIOWY(7) ,NFTCFLOW(7) ,TWMASS(7) 

REAL CCNa™,CCN04AX,CMASSMAX,CMAS ,CSMASS(23,45) 

REAL CFLOWX(23,45) ,CFLOWY(23,45) ,CFLOWXMAX,CFD0WXMm,aTXWYMAX 

REAL DRV0IMAX,DRV0IKLN,DRV0ISUM(7) ,MST0RSUM(7) ,TEMP 

INTEGER*2 NX,NY,I, J,:QNIC,II, JJ,KK,LL,MM,NN,YY(7) ,XX(2) ,STRT,STP 

INTEGER*2 IYR, IMON, IDAY, IHR, IMIN, ISEC, I100TH, IOUT 

CHARACTER*2 MONTH, DAY, YEAR 

CHARACTER*8 DATE,CHEMFILE 

CHARACTER* 12 TNFTIE1 , INFIIE2 , INFIIE3 , INFILE4 , INFIIE6 , OUTFIIE 
CHARACTER* 12 FLUXFILE, COMMENT (14) 
CCCCCC 

C INITIALIZE VARIABLES 
C PI = PI 

C COND(J) = SAT. HYDRAULIC CONDUCTIVITY (m/day) AT NODE J 

C NODECOND(J) = SAT. HYDRUALIC COND. AT SPECIFIED NODES J 

C FORD = SOIL POROSITY (cm/cm) 

C KD = PESTICIDE DISTRIBUTION COEFF. (cm**3/g) 

C BULKDEN = SOIL BULK DENSITY (g/cm**3) 

CCCCCC 

PI = 3.141592654 

PORO = 0.36 

KD = 0.0489 

BULKDEN = 1.59 

INFILE1 = 'IMPERMBA.GRD' 

INFIIE3 = ' SURFACEA. GRD ' 

INFILE2 = 'INFIIE .DAT' 

CCCCCC 

C FILE INFILE.DAT CONTAINS MONTH, DAY, YEAR VALUES FOR GENERATING 
C FILENAMES AND DATE 

C FILE IMPERMBA.GRD CONTAINS GRIDDED ELEVATIONS OF IMPERMEABLE LAYER 
C FILE SURFACEA. GRD CONTAINS GRIDDED ELEVATIONS OF SOIL SURFACE 
CCCCCC 



177 



178 



Figure C.l Continued. 

OPEN (UNIT=1 , FILEKQNFILE1 , STATUS= ' OLD ' ) 
OPEN (UNTT=3 , FILE=LNFILE3 , STATUS= 1 OLD 1 ) 
OPEN (UNIT=2 , FLLE=LNFLLE2 , STATUS= 1 OLD 1 ) 
CCCCCC 

C READ IN IMPERMEABLE LAYER ELEVATION DATA 
C NX, NY ARE # OF X AND Y VALUES IN GRID 

C XMLN, YMLN AND XMAX, YMAX ARE MLN AND MAX VALUES ON RESPECTIVE AXIS 
CCCCCC 

READ(1,*) 
READ(1,*) NX, NY 
READ(1, *) XMLN, XMAX 
READ(1,*) YMLN, YMAX 
READ(1,*) IMPMLN, LMRMAX 
DO 5, J= 1,NY 

READ(1,*) (IMP(I,J), 1=1, NX) 

READ(1,*) 
5 CONTINUE 
CCCOCOC 

C CALCULATE SPACING OF GRID POINTS IN X AND Y DIRECTIONS 
C DELX = DISTANCE BETWEEN NODES ON X-AXIS 
C DELY = DISTANCE BETWEEN NODES ON Y-AXIS 
CCCCCCC 

DELX = (XMAX-XMLN) / (NX-1) 

DELY = (YMAX-YMLN) / (NY-1) 

CCCCCC 

C READ LN SOIL SURFACE ELEVATION DATA 
CCCCCC 

READ(3,*) 
READ(3,*) 
READ(3,*) 
READ(3,*) 

READ(3,*) SURFMIN,SURFMAX 
DO 10, J= 1,NY 

READ(3,*) (SURF(I,J), 1=1, NX) 

READ(3,*) 
10 CONTINUE 
CCCCCC 
C READ LN : 

C OUTPUT INCREMENT VALUE, PRINT EVERY INC NODES LN X AND Y 
C CONDUCTIVITY (m/day) AT Y-AXIS BOUNDARY NODES 

C BOUNDARY NODE NUMBERS ON X-AXIS FOR SUBAREA BOUNDARIES, 2 VALUES 
C BOUNDARY NODE NUMBERS ON Y-AXIS FOR SUBAREA BOUNDARIES, 5 VALUES 
C SUMMARY FILE NAME 

C . .AND FOR 28 FILES READ THE FOLLOWING. . . 
C MONTH, DAY, YEAR AND CHEMICAL FILENAME 
CCCCCC 

C CREATE INPUT AND OUTPUT FILENAMES BY ADDING APPROPRIATE EXTENSIONS 
CCCCCC 

READ(2 , 102) INC, IOUT 

READ(2,104) (NODECOND(I) ,1=1,7) 



179 



Figure C.l Continued. 

READ (2, 102) XX (1) ,XX(2) 
KEAD(2,102) (YY(I) ,1=1,7) 
READ ( 2 , 108 ) FIUXFIIE 
READ(2,933) (COMMENT(I) ,1=1,14) 
CCCCCC 

C INTERPOLATE CONDUCTIVITIES AT EACH Y NODE 
CCCCCC 

DO 621, I = 1,45 
DO 621, J = 1,6 

IF(I.GE.YY(J) .AND.I.LE.YY(J+1)) THEN 
COND(I) = NODEOOND(J) + FLOAT (I-YY (J) ) * (NODECOND ( J+l) - 
# NODECOND (J)) / (YY(J+1) - YY(J)) 
ELSEIF(I.LE.YY(1)) THEN 

COND(I) = NODECOND (1) 
ELSELF(I.GE.YY(7) ) THEN 

OOND(I) = NODECOND(7) 
ENDIF 
621 CONTINUE 
CCCCCC 

C OPEN OUTPUT FILE FOR TRANSFER OF DATA TO WEEKLY FLUX QujCULATION 

C PROGRAM 

CCCCCC 

OPEN (UNLT=7 , FILE=FLUXFILE, STATUS= 'UNKNOWN 1 ) 
WRTTE(7,934) (COMMENT(I) ,1=1,14) 
CALL GET1'1M(IHR, LMLN, ISEC, I100TH) 
CALL GETDAT (IYR, IMON, IDAY) 

WRITE(7,171) IMON, IDAY, IYR, IHR, BUN, ISEC, I100TH 
WRITE (7, 195) (NODECOND(I) ,1=1,7) 
WRITE(7,196)PORO 
WRITE (7, 921) KD 
WRITE (7 , 922 ) BULKDEN 
WRTTE(7,731)XX(1) ,XX(2) 
WRITE(7,732) (YY(I) ,1=1,7) 
CCCCCC 

C BEGIN MAIN LOOP TO PERFORM CALCULATIONS FOR EACH OF 28 OBSERVATION 

C PERIODS 

CCCCCC 

30 DO 15, K=l,50 

READ(2,100,END=20) MONTH, DAY, YEAR, CHEMFILE 
DATE = MONTH // •/ 1 // DAY // •/ 1 // YEAR 
LNFILE4 = MONTH // •-' // DAY // 'ELV // 1 .GRD' 
LNFILE6 = CHEMFILE // ' .GRD' 

OUTFTLE = MONTH // •-' // DAY // 'WTD' // '.SUM' 
WRITE(*,*) INFIIE4 , LNFILE6 , OUTFILE 
CCCCOOC 

C OPEN INPUT AND OUTPUT FILES 
CCCCCCC 

OPEN (UNIT=4 , FILE=INFILE4 , STATUS= ' OLD 1 ) 
OPEN (UNTT=5 , FIIE^OUTFIIE , STATUS= ' UNKNOWN ' ) 
OPEN (UNTT=6 , FILE=LNFILE6 , STATUS= 1 OLD 1 ) 



180 



Figure C.l Continued. 
CCCCCCC 

C READ IN VAIUES FRCM INPUT FILES 

C WIMIN AND WIMAX ARE MLN AND MAX VALUES OF WT ELEVATION 

C CONCMLN AND CONCMAX ARE MIN AND MAX CHEM CONCENTRATION VALUES 

CCCCCCC 

READ(4,*) 

READ(4,*) 

READ(4,*) 

READ(4,*) 

READ(4,*) WIMIN, WIMAX 

OCCCOC 

READ(6,*) 
READ(6,*) 
READ(6,*) 
READ(6,*) 

READ(6,*) CONCMIN, CONCMAX 

CCCCOC 

C ZERO OUT OLD VALUES AT THE NODES AND READ IN NEW VALUES OF WT(I,J) 

C AND CONC(I,J) 

CCCCOC 

DO 35 1=1,23 
DO 35 J=l,45 
WT(I,J) = 0.0 
WTD(I,J) = 0.0 
CONC(I,J) =0.0 
SLPX(I,J) = 0.0 
SLFY(I,J) = 0.0 
VELX(I,J) = 0.0 
VELY(I,J) = 0.0 
FLOWX(I,J) = 0.0 
FLOWY(I,J) = 0.0 
CFLOWX(I,J) = 0.0 
CFLOWY(I,J) = 0.0 
MAGNITUDE (I, J) = 0.0 
DEFECTION (I, J) =0.0 
DRVOL(I,J) = 0.0 
MAXSTOR(I,J) = 0.0 
35 CONTINUE 

DO 45, J= 1,NY 

READ(4,*) (WT(I,J), 1=1, NX) 

READ(4,*) 

READ(6, *) (CONC(I,J), 1=1, NX) 

READ(6,*) 
45 CONTINUE 
CCCOCC 

C NOW COMEUTE SLOPE AND MAGNITUDE AT INTERIOR NODES USING CENTERED 
C DIFFERENCE TECHNIQUES 

C FIRST COMPUTE SLOPE INDIVIDUALLY IN X AND Y DIRECTIONS 
C MINUS SIGN GIVES ACTUAL SLOPE AND NOT GRADIENT 
CCCCOC 



181 

Figure C.l Continued. 

DO 50, J= 2, (NY-1) 
DO 50, 1= 2, (NX-1) 

SLPX(I,J) = -( WT(I+1,J) - WT(I-1,J) ) / ( 2.0*DELX ) 

SLPY(I,J) = -( WT(I,J+1) - WT(I,J-1) ) / ( 2.0*DELY ) 

CCCCCCC 

C COMPUTE MAGNITUDE OF RESULTANT SLOPE 

cxxxrcc 

MAGNITUDE (I, J) = SQRT ( SLPX(I,J)**2 + SLPY(I,J)**2 ) 

CCCCCCC 

C COMPUTE DIRECTION OF RESULTANT SLOPE 

C ANGLE IS MEASURED wrt +X AXES, ie ANGLE = 90 = +Y AXIS 
CCCCCCC 

ANGLE(I, J) = ATAN2 ( SLPY(I,J) , SLPX(I,J) ) * 180.0 / PI 
IF (ANGLE(I,J) .LT.0.0) THEN 

DIRECTION (I, J) = 360.0 + ANGLE(I,J) 
ELSE 

DLRECTION(I,J) = ANGLE (I, J) 
ENDIF 

CCOCCC 

C CALCULATE WATER TABLE DEPTH (cm) AT EACH NODE 
CCCCCC 

WTD(L,J) = 100.0 * (WT(I,J) - IMP(I,J)) 
CCCCCC 

C CALCULATE DRAINED VOLUME (m**3) AT EACH NODE 
CCCCCC 

WTDEPTH = SURF (I, J) - WT(I,J) 

DRVOL(I,J) = VOITRAIN (WTDEPTH) * DELX * DELY / 100.0 
CCCCCC 

C CALCULATE MAXIMUM STORAGE VOLUME (m**3) AT EACH NODE 
CCCCCC 

MAXSTOR(I,J) = (SURF(I,J) - IMP(I,J)) * PORO * DELX * DELY 
CCCCCC 

C CALCULATE VELOCITY, AND FLOWRATE IN X AND Y DIRECTIONS 
C AT EACH NODE. 

C VELX(I,J) = VELOCITY IN X-DIRECTION (m/day) 

C FLOWX(I,J) = VOLUMETRIC FLOWRATE IN X-DIRECTION (m**3/day) 

CCCCCC 

VELX(I,J) = COND(J) * SLPX(I,J) / PORO 
VELY(I,J) = COND(J) * SLPY(I,J) / PORO 

FLOWX(I,J) = COND(J) * SLRX(I,J) * DELX * WTD(I,J) / 100.0 
FLOWY(I,J) = COND(J) * SLPY(I,J) * DELY * WTD(I,J) / 100.0 
CFLOWX(I,J) = CONC(I,J) * FLOWX(I,J) 
CFLOWY(I,J) = CONC(I,J) * FLOWY(I,J) 
CCCCCC 

C CALCULATE MASS OF WATER AND CHEMICAL AT EACH NODE 

C WMASS = WATER STORAGE IN SATURATED ZONE (m**3) IN NODAL AREA 

C CWMASS = MASS OF CHEMICAL IN WATER (mg) IN NODAL AREA 

C CSMASS = MASS OF CHEMICAL ON SOIL (mg) IN NODAL AREA 

C CMASS = TOTAL MASS OF CHEMICAL (mg) " » 

CCCCCC 



Figure C.l Continued. 



WMASS(I,J) = WTD(I,J) * PORQ * DELX * DELY / 100.0 
CWMASS(I,J) = WMASS(I,J) * C0NC(I,J) 
CSMASS(I,J) = CWMASS(I,J) * KD * BULKDEN / PORO 
CMASS(I,J) = CWMASS(I,J) + CSMASS(I,J) 

CCCCOC 

C CAITUIATE MAX AND MIN VALUES OF SELECTED PARAMETERS 
CCCCCC 

IF(I.EQ.2.AND.J.EQ.2) THEN 
WTDMAX = WTD(2,2) 
WTDMLN = WTD(2,2) 
SLPXMAX = SLPX(2,2) 
SLPXMLN = SLPX(2,2) 
SLPYMAX = SLPY(2,2) 
SLPYMXN = SLPY(2,2) 
MAGMAX = MAGNITUDE (2, 2) 
MAGMXN = MAGNITUDE (2, 2) 
FLDWXMAX = FLOWX(2,2) 
FLOWXMIN = FIOWX(2,2) 
FLOWYMAX = FLOWY(2,2) 
FLOWYMTN = FLOWY(2,2) 
CFLOWXMAX = CFIOWX(2,2) 
CFLOWXMIN = CFLOWX(2,2) 
CFLOWYMAX = CFLDWY(2,2) 
CFLDWYMIN = CFLOWY(2,2) 
CMASSMAX = CMASS(2,2) 
CMASSMIN = CMASS(2,2) 
ENDIF 

IF(WTD(I,J) .GT. WTDMAX) WTDMAX = WTD(I,J) 
IF(WTD(I,J) .LT.WTDMIN) WTDMLN = WTD(I,J) 
IF(SLPX(L,J) .LT. SLPXMLN) SLPXMLN = SLPX(I,J) 
TF(SLPX(I,J) .GT. SLPXMAX) SLPXMAX = SLPX(I,J) 
LF(SLPY(L,J) .LT.SLPYMLN) SLPYMLN = SLPY(L,J) 
IF(SLPY(L,J) .GT. SLPYMAX) SLPYMAX = SLPY(L, J) 
LF (MAGNITUDE (L, J) .GT. MAGMAX) MAGMAX = MAGNITUDE (I, J) 
LF (MAGNITUDE (I, J) .LT.MAGMLN) MAGMLN = MAGNITUDE (I, J) 
IF(FLOWX(L,J) .GT.FIDWXMAX) FLDWXMAX = FIOWX(L,J) 
LF(FLOWX(L,J) .LT.FLOWXMLN) FLDWXMLN = FLOWX(L,J) 
LF(FLOWY(L, J) .GT. FLOWYMAX) FLOWYMAX = FLOWY(L,J) 
LF(FLOW¥(L,J) .Lr.FIOWYMLN) FLOWYMLN = FLOWY(L,J) 
LF(CFLOWX(L,J) .GT.CFIOWXMAX) CFLOWXMAX = CFIOWX(L,J) 
LF(CFLOWX(L,J) .LT.CFLOWXMLN) CFLOWXMLN = CFLOWX(L,J) 
IF(CFIDWY(L, J) .GT. CFLOWYMAX) CFLOWYMAX = CFLOWY(L,J) 
LF(CFLOWY(L, J) .IT.CFLOWYMTN) CFLOWYMLN = CFLOWY(I,J) 
LF(CMASS (I, J) .GT. CMASSMAX) CMASSMAX = CMASS(L, J) 
IF(CMASS(L, J) .LT.CMASSMLN) CMASSMIN = CMASS(I,J) 

CCCCCC 

C END NODAL CALCULATION LOOPS 
CCCCCC 

50 CONTINUE 



183 



Figure C.l Continued. 

cccccc 

C BEGIN CMOILATIONS FOR SUB-AREAS AND TOTAL CONTROL AREA 
CCCCCC 

C CALCULATE MASS FLOWS ACCROSS BOUNDARIES OF THE SIX AREAS 

C BFLOWXl(I) = FLOW ACCROSS PLANES WITH X = XX (1) , Y = YY(I) TO YY(I+1) 

C BFLOWX2(I) = FLOW ACCROSS PIANES WITH X = XX(2) , Y = YY(I) TO YY(I+1) 

C BFLOWY(I) = FLOW ACCROSS PIANES WITH Y = YY(I) , X = XX (1) TO XX(2) 

CCCCCC 

C FIRST FLOWS ACCROSS Y PIANES 
CCCCCC 

DO 200, KK = 1,7 
BFLOWY(KK) = 0.0 
BCFLOWY(KK) = 0.0 
DO 200, LL = XX (1) ,XX(2) 

IF(LL.EQ.XX(1) .OR.LL.EQ.XX(2)) THEN 
C **HALF AREA CONTRIBUTING ON BOUNDARIES** 

BFLOWY(KK) = BFIOWY(KK) + FLOWY(LL, YY(KK) )/2.0 
BCFIOWY(KK) = BCFIOWY(KK) + CFLOWY (LL, YY (KK) ) /2 . 
ELSE 

BFLOWY(KK) = BFLOWY(KK) + FLOWY (LL, YY (KK) ) 

BCFIOWY(KK) = BCFIOWY(KK) + CFLOWY (LL, YY (KK) ) 
ENDIF 
200 CONTINUE 
CCCCCC 

C NOW FLOWS ACCROSS X-PIANES 
CCCCCC 

DO 210, KK = 1,7 
BFLOWXl(KK) = 0.0 
BFLOWX2(KK) = 0.0 
BCFLOWXl(KK) = 0.0 
BCFIDWX2 (KK) = 0.0 
STRT = YY(KK) 
STP = YY(KK+1) 
IF(KK.EQ.7) THEN 
STRT - YY(1) 
STP = YY(7) 
ENDIF 
DO 210 LL = STRT, STP 

IF(LL.EQ.STRT.OR.LL.EQ.STP) THEN 
C **HALF AREA CONTRIBUTING ON BOUNDARIES** 

BFIDWX1 (KK) = BFLOWXl(KK) + FLOWX(XX(l) ,LL)/2.0 
BFLOWX2(KK) = BFLOWX2(KK) + FIOWX(XX(2) ,LL)/2.0 
BCFLOWXl(KK) = BCFLOWXl(KK) + CFLOWX(XX(l) ,LL)/2.0 
BCFIOWX2(KK) = BCFLOWX2(KK) + CFLOWX(XX(2) ,LL)/2.0 
ELSE 

BFLOWXl(KK) = BFIOWXl(KK) + FLOWX(XX(l) ,LL) 
BFLOWX2(KK) = BFLOWX2(KK) + FLOWX(XX(2) ,LL) 
BCFLOWXl(KK) = BCFIOWXl(KK) + CFLOWX(XX(l) ,LL) 
BCFLOWX2(KK) = BCFLOWX2(KK) + CFLOWX(XX(2) ,LL) 
ENDIF 



184 



Figure C.l Continued. 

210 CONTINUE 
CCCCCC 

C CAICUTATE NET MASS FDJXES TN EACH OF THE SIX SUB-AREAS AND FOR TOTAL 

C AREA. 

CCCCCC 

DO 220, MM = 1,6 

NETFLOW (MM) =BFIDWY (MM) + BFLOWXl(MM) - BFLOWX2(MM) - BFLOWY(MM+l) 
NETCFLOW (MM) =BCFLOWY (MM) + BCFIOWXl(MM) - BCFLOWX2(MM) - 
# BCFLOWY(MM+l) 
220 CONTINUE 

NETFLOW(7) = BFIOWY(l) + BFIDWX1(7) - BFLOWX2(7) - BFLOWY(7) 
NETCFLOW(7) = BCFIOWY(l) + BCFIDWX1(7) - BCFLOWX2(7) - BCFLOWY(7) 
CCCCCCC 

C COMPUTE MEAN WTD, TOTAL WATER AND CHEMICAL MASSES IN EACH AREA 
CCCCCCC 

DO 310, KK = 1,7 
TWMASS(KK) = 0.0 
TCMASS(KK) = 0.0 
DRVOLSUM(KK) =0.0 
MSTORSUM(KK) =0.0 
MEANWTD(KK) =0.0 
AREA (KK) =0.0 
STRT = YY(KK) 
STP = YY(KK+1) 
IF(KK.EQ.7) THEN 
STRT = YY(1) 
STP = YY(7) 
ENDIF 
DO 300, J = STRT, STP 

DO 300, I = XX(1) ,XX(2) 
IF(I.EQ.XX(1) .OR.I.EQ.XX(2)) THEN 
IF(J.EQ.STRT.OR.J.EQ.STP) THEN 
C ** QUARTER AREA CONTRIBUTION IN CORNERS ** 

TWMASS(KK) = TWMASS(KK) + WMASS(I, J)/4.0 
TCMASS(KK) = TCMASS(KK) + CMASS(I, J)/4.0 
MSTORSUM(KK) = MSTORSUM(KK) + MAXSTOR(I, J)/4.0 
DRVOLSUM(KK) = DRVOLSUM(KK) + DRVOL(I, J)/4.0 
AREA(KK) = AREA(KK) + DELX * DELY / 4.0 
MEANWTD(KK) = MEANWTD(KK) + WTD(I, J) *DELX*DELY/4.0 
C WRITE(5,951)KK,I,J,TWMASS(KK) ,WMASS(I,J) 

C951 FORMAT(1X,3I3,2(4X,F10.5) , 1 QUARTER AREA' ) 

ELSE 

C **HALF AREA CONTRIBUTING ON BOUNDARIES** 

TWMASS(KK) = TWMASS(KK) + WMASS(I, J)/2.0 
TCMASS(KK) = TCMASS(KK) + CMASS(I,J)/2.0 
MSTORSUM(KK) = MSTORSUM(KK) + MAXSTOR(I, J)/2.0 
DRVOLSUM(KK) = DRVOLSUM(KK) + DRVOL(I, J)/2.0 
AREA(KK) = AREA(KK) + DELX * DELY / 2.0 
MEANWTD (KK) = MEANWTD (KK) + WTD ( I , J) *DELX*DELY/2 . 

C WRITE(5,952)KK,I,J,TWMASS(KK) ,WMASS(I,J) 



185 



Figure C.l Continued. 

C952 FORMAT(1X,3I3,2(4X,F10.5) , • HALF AREA') 

ENDIF 

ELSEIF(J.BQ.STFT.OR.J.BQ.STP) THEN 
IF(I.EQ.XX(1) .OR.I.EQ.XX(2)) THEN 
GO TO 300 
ELSE 

TWMASS(KK) = TWMASS(KK) + WMASS(I, J)/2.0 
TCMASS(KK) = TCMASS(KK) + OMASS(I, J)/2.0 
MSTORSUM(KK) = MSTORSUM(KK) + MAXSTOR(I, J)/2.0 
DRVOISUM(KK) = DRVOLSUM(KK) + DRVOL(I, J)/2.0 
AREA(KK) = AREA(KK) + DELX * DELY / 2.0 
MEANWTD(KK) = MEANWTD(KK) + WTD(I,J) *DELX*DELY/2.0 
C WRTTE(5,952)KK,I,J,TWMASS(KK) ,WMASS(I,J) 

ENDIF 
ELSE 

TWMASS(KK) = TWMASS(KK) + WMASS(I,J) 
TCMASS(KK) = TCMASS(KK) + CMASS(I,J) 
MSTORSUM(KK) = MSTORSUM(KK) + MAXSTOR(I,J) 
DRVOLSUM(KK) = DRVOLSUM(KK) + DRVOL(I,J) 
AREA(KK) = AREA(KK) + DELX * DELY 
MEANWTD(KK) = MEANWTD(KK) + WTD(I, J) *DELX*DELY 



C WRITE(5,953)KK,I,J,TWMASS(KK) ,WMASS(I,J) 

C953 FORMAT(1X,3I3,2(4X,F10.5) , 1 FULL AREA 1 ) 

ENDIF 
300 CONTINUE 

MEANWTD(KK) = MEANWTD(KK) / AREA(KK) 
310 CONTINUE 



OCCCCC 

C WRITE NODAL VALUES TO OUTEUT FILE 

C XPOS AND YPOS ARE THE X AND Y POSITIONS IN THE FIELD IN METERS 

C ROW AND COL CORRESPOND TO THE STATION ID LOCATIONS 

OCCCCC 

IF(IOUT.GT.O) THEN 
WRTTE(5,110) DATE 
C WRITE(5, 623) (COND(I) ,1=1,45) 
C623 FORMAT (10F7. 2) 
WRITE (5, 120) 

DO 60, J= 2, (NY-1), INC 
DO 60, 1= 2, (NX-1) , INC 
IF(J.EQ.14.AND.INC.GT.1.AND.I.EQ.2) THEN 
JJ = 11 

DO 80 II = 10,18,4 
XPOS = FLOAT (II-l) * DELX + XMLN 
YPOS = FLOAT (JJ-1) * DELY + YMIN 
ROW = 6.0 + FLOAT (II-2) / 4.0 
COL = 7.0 + FLOAT (JJ-2) / 4.0 
WRITE(5, 130)11, JJ, ROW, COL, XPOS, YPOS, SURF(II,JJ) ,IMP(II,JJ) , 

# WT(II,JJ) ,WID(II, JJ) ,SLPX(II,JJ) ,SLPY(II, JJ) ,MAGNITUDE(II, JJ) , 

# DIRECTION ( II, JJ) 
80 CONTINUE 



186 



Figure C.l Continued. 
ENDIF 

XPOS = FLOAT(I-l) * DELX + XMLN 
YPOS = FLOAT (J-l) * DELY + YKEN 
ROW = 6.0 + FLOAT (1-2) / 4.0 
OOL = 7.0 + FLOAT (J-2) / 4.0 

WRITE(5,130)I,J,ROW,COL,XF0S,YFOS,SURF(I,J) ,IMP(I,J) , 

# WT(I,J) ,WTD(I,J) ,SLPX(I,J) ,SIFY(I,J) ,MAGNITUDE(I, J) , 

# DIRECTION (I, J) 
60 CONTINUE 

WRITE(5,140) 
WRITE(5,431) 
WRTTE(5,432) 
WRTTE(5,433) 

DO 320, J= 2, (NY-1) , INC 
DO 320, 1= 2, (NX-1) , INC 
IF(J.EQ.14.AND.INC.GT.1.AND.I.EQ.2) THEN 
JJ = 11 

DO 380 II = 10,18,4 

WRITE(5,142)II,JJ,MAXSTOR(II,JJ) ,DRVOL(II, JJ) ,WMASS(II, JJ) , 

# CONC(II,JJ) ,CWMASS(II,JJ) ,CSMASS(II,JJ) ,CMASS(II, JJ) , 

# FLOWX(II,JJ) ,VELX(II,JJ) ,CFIOWX(II,JJ) ,FIOWY(II, JJ) , 

# VELY(II,JJ) ,CFIOWY(II,JJ) 
380 CONTINUE 

ENDIF 

WRITE(5,142)I,J,MAXSTOR(I,J) ,DRVOL(I,J) ,WMASS(I,J) ,OONC(I,J) , 

# CWMASS(I,J) ,CSMASS(I,J) ,CMASS(I,J) ,FIOWX(I,J) ,VELX(I,J) , 

# CFIOWX(I,J) ,FIOWY(I,J) ,VELY(I,J) ,CFLOWY(I,J) 
320 CONTINUE 

CCCCCC 

C WRITE BOUNDARY FLUX AND STORAGE VALUES 
CCCCCC 



WRITE 


15 


140) 




WRITE 


|5 


180) 




WRITE 


;5 


448) 


(AREA(I) ,1=1,7) 


WRITE 


[5 


181) 


(BFIOWXl(I) ,1=1,7) 


WRITE 


[5 


182) 


(BFLOWX2(I),I=l,7) 


WRITE 


[5 


183) 


(BFLOWY(I) ,1=1,6) ,BFIOWY(l) 


WRITE 


[5 


421) 


(BFLOWY(I) ,1=2,7) ,BFIOWY(7) 


WRITE 


[5 


187) 


(NETFIOW(I) ,1=1,7) 


WRITE 


[5 


412) 


(MSTORSUM(I) ,1=1,7) 


WRITE 


[5 


411) 


(DRVOLSUM(I) ,1=1,7) 


WRITE 


[5 


191) 


(TWMASS(I) ,1=1,7) 


WRITE 


[5 


447) 


(MEANWTD(I) ,1=1,7) 


WRITE 


[5, 


184) 


(BCFIOWXl(I) ,1=1,7) 


WRITE 


'5, 


185) 


(BCFLOWX2(I) ,1=1,7) 


WRITE 


'5, 


186) 


(BCFLOWY(I) ,1=1,6) ,BCFIOWY(l) 


WRITE 


'5, 


422) 


(BCFIOWY(I) ,1=2,7) ,BCFLOWY(7) 


WRITE! 


'5, 


188) 


(NETCFIOW(I) ,1=1,7) 


WRITE 


'5, 


192) 


(TCMASS(I) ,1=1,7) 



Figure C.l Continued. 
CCCCCC 

C WRITE MAXIMUMS AND MTNIMUMS 
CCCCCC 

WRITE (5, 149) 

WRITE(5,150)SUREMIN,SURFMAX 
WRITE(5,151)TMPMIN,IMPMAX 
WRITE(5, 152)WTMIN,WIMAX 
WRITE(5, 153)WTOyiIN / WrDMAX 
WRrTE(5, 154) SLPXMIN^LPXMAX 
WRITE ( 5 , 155) SLPYMIN, SLPYMAX 
WRITE(5, 156)MAGMIN / MAGMAX 
WRITE (5 ,157) FLOWXMTN, FLOWXMAX 
WRITE (5 , 158) FLOWYMIN, FLOWYMAX 
WRITE (5 , 163 ) CFIDWXMIN, CTTOWXMAX 
WRITE(5,164)CIFIOWYMIN,CFL0WYMAX 
WRTTE(5, 165) CMASSMIN, CMASSMAX 
WRITE(5,166)C0NCMIN,CONCMAX 
CCCCCC 

C POST FILENAMES USED IN THIS ANALYSIS 
CCCCCC 

WRITE (5, 160) 

WRITE (5, 161) LNFILE1 , OUTFILE 
WRITE (5,162) LNFILE2 , LNFILE3 , LNFILE4 , LNFILE6 
WRTTE(5,195) (NODECOND(I) ,1=1,7) 
WRITE(5,196)PORO 
WRITE(5,921)KD 
WRITE ( 5 , 922 ) BULKDEN 
WRITE (5, 731) XX (1) ,XX(2) 
WRTTE(5,732) (YY(I) ,1=1,7) 
WRTTE(5,934) (COMMENT(I) ,1=1,14) 
CCCCCC 

C CALL DATE AND TIME TO MARK OUTPUT FOR LATER REFERENCE 
CCCCCC 

CALL GETriM(IHR, TMLN, ISEC, I100TH) 
CALL GETDAT(IYR,IMON,IDAY) 

WRITE (5, 170) IMON, IDAY, IYR, THR, IMLN, ISEC, I100TH 
ENDIF 
CCCCCC 

C WRITE TO FLUX SUMMARY OUTPUT FILE 
CCCCCC 

WRITE (7, 452) (AREA(I) ,1=1,7) , DATE 
WRTTE(7,401) (NETFLOW(I) ,1=1,7) ,DATE 
WRTTE(7/402) (TWMASS(I) ,1=1,7) ,DATE 
WRrTE(7,405) (DRVOLSUM(I) ,1=1,7) ,DATE 
WRITE(7,406) (MSTORSUM(I) ,1=1,7) ,DATE 
WRITE (7, 451) (MEANWTD(I) ,1=1,7) , DATE 
WRITE(7,403) (NETCFLOW(I) ,1=1,7) ,DATE 
WRTTE(7,404) (TCMASS(I) ,1=1,7) ,DATE 



188 



Figure C.l Continued. 
CCCCCC 

C CLOSE FILES 
CCCCCC 

CLOSE (UNIT=4 , STATUS= ' KEEP ' ) 
CIDSE (UNIT=5 , STATUS= 1 KEEP 1 ) 
CLOSE ( UNIT=6 , STATUS= ' KEEP ' ) 
CCCCCC 

C LOOP BACK FOR NEXT FILE 
CCCCCC 

15 CONTINUE 
CCCCCC 

C STOP FOR NORMAL TEFMLNATION 
CCCCCC 

20 STOP 1 END OF FILE ON INFILE.DAT 1 

CCCCCC 

C FORMAT STATEMENTS 
CCCCCC 

100 FORMAT ( 3 (A2, IX) ,A8) 
102 FOFMAT(10I3) 
104 FORMAT (7F10. 2) 
106 FORMAT (II) 
108 FORMAT (A12) 

110 FORMAT ( • ' ,/,45X, 'GOPHER RIDGE WATER TABLE DATA* ,/,52X, 'DATE = 1 , 

# A8,//,2X,118('-')) 

120 FORMAT (' ' ,2X, 'NODE STATION ID LOCATION (m) ELEVATION ' 

#•(10) WT DEPTH WATER TABLE SLOPE (m/m) 

# 'DIRECTION', /,4X, 'I J',17X, 'X',7X, 'Y',5X, 'SURF',4X, 'IMP',6X, 'WT' , 

# 6X, ' (cm) ',10X, «X',9X, 'Y',8X, 'MAGNITUDE' ,2X, ' (degrees) ' ,/, 

# 2X, 118 ('-')) 

130 FORMAT( ' ' ,1X,I2,1X,I2,2X,F5.2, '-' ,F5.2,2X,F5.2,2X,F6.2, 

# 3(2X,F6.3),3X,F5.1,4X,3(E10.4,2X) ,3X,F4.0,4(F7.3,1X) ) 
140 FORMAT ( ' ',118 ('-')) 

431 FORMAT (IX, 126 ('-') ,/,3X, 'I J MAXIMUM DRAINED SATURATED ', 

#' CHEMICAL CHEMICAL STORAGE ' ,2X,8 ('-'), 'X DIRECTION' ,7 ('-•) , 
#3X,6( '-' ) , 'Y DTRECnON', 8 ('-')) 

432 FORMAT (10X, 'STORAGE VOLUME STORAGE CONC. ',6X, 

#' WATER SOIL TOTAL FLOWRATE VELOCITY CHEMICAL FLOWRATE ' , 
#• VELOCrTY CHEMICAL') 

433 FOFMAT(9X, ' (m**3) (m**3) (m**3) (ppb) ',3(' (mg) ') 
#, 1 (m**3/day) ',1X, ' (m/day) (mg/day) (m**3/day) (m/day) ', 

# • (mg/day) • ,/,lX,126( •-•) ) 

142 PORMAT(2X,I2,2X,I2,3(F9.4) ,F9.2,9(F9.3) ) 

149 FORMAT ( 1 1 ,3X,53 ('-') ,/,5X, 'PARAMETER' , 18X, 

# 'MINIMUM', 7X, 'MAXIMUM' ,/,5X,53 ( •-• ) ) 

150 FORMAT ( ' ' ,4X, 'SOIL SURFACE ELEVATION' ,5X,F5. 2, 9X,F5. 2) 

151 FORMAT ( 1 \4X, 'IMP. LAYER ELEVATION' ,5X,F5. 2, 9X,F5. 2) 

152 FORMAT( ' ' ,4X, 'WATER TABLE ELEVATION' ,5X,F5. 2, 9X,F5. 2) 

153 FOEMAT^ 1 1 ,4X, 'WATER TABLE DEPTH' ,9X,F5.1,9X,F5.1) 

154 FORMAT ( 1 ',4X, 'SLOPE IN X-DTRECTION 1 , 3X, E10 . 4 , 5X, E10 . 4 ) 

155 FORMAT ( ' ',4X, 'SLOPE IN Y-DTRECTION 1 , 3X, E10 . 4 , 5X, E10 . 4 ) 



189 



Figure C.l Continued. 



FORMAT 
FORMAT 
FORMAT 
FORMAT 

# 5X,11 
FORMAT 
FORMAT 

# 'UNIT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 

# 12.2, 
FORMAT 

# 12.2, 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 
FORMAT 



i _ 



,4X, 'SLOPE MAGNITUDE ',9X,E10. 4, 5X,E10. 4) 
,4X, 'FIOW IN X-DIRECITON',3X,E10.4,5X,E10.4) 
,4X, 'FIOW IN Y-DIRECITON',3X,E10.4,5X,E1C).4) 
,4X,53 (*-*) ,//,5X, 'INPOT FILES', 14X, 'OUTPUT FILE',/, 
),14X,11('-')) 

,4X, 'UNIT 1 = ',A12,4X, 'UNIT 5 = ' ,A12) 
,4X, 'UNIT 2 = 1 ,A12,/,5X, 'UNIT 3 = ',A12,/,5X, 
= ',A12,/,5X, 'UNIT 6 = \A12) 
,4X, 'CHEM FIOW IN X-DIR' ,3X,E10.4,5X,E10.4) 
,4X, 'CHEM FIOW IN Y-DIR' ,3X,E10.4,5X,E10.4) 
,4X, 'CHEMICAL MASS STORAGE' ,3X,E10.4,5X,E10.4) 
,4X, 'CHEMICAL CONCENrRATION 1 ,3X,E10.4,5X,E10.4) 
/,4X, 'PROGRAM RUN ON » ,12.2, '/' ,12.2, •/' ,14, • AT ', 
12.2, '.',12.2, '.'12.2) 

'PROGRAM RUN ON ',12.2, '/', 12.2, '/', 14, ' AT', 
12.2, '.',12.2, '.'12.2) 

2X,110('-')) 
2X)) 



4X 
i 



2X, 1 BOUNDARY FLUXES',/ 
\7(F10.5 
',7(F10.5 

' BFLOWX2(I) 
' BFLOWYL(I) 
• BFLOWY2(I) 



' AREA(I) 
■ BFIOWXl(I) 



',7(F10.5 
',7(F10.5 
',7(F10.5 



/,' BCFIOWXl(I) ',7(F10.5,2X)) 



' BCFIOWX2(I) ',7(F10.5 

• BCFLOWYl(I) ',7(F10.5 

• BCFLOWY2(I) ',7(F10.5 
' NETFIOW(I) ',7(F10.5 
' NETCFLOW(I) ',7(F10.5 

• TWMASS(I) \7(F10.5 

• MEANWTD(I) 
' TCMASS(I) 
1 DRVOL(I) 
' MAXSTOR(I) 

HYD COND. 



POROSITY 
NETFLOW(I) 
/,' AREA(I) 
TWMASS(I) 
NETCFIOW(I) 
TCMASS(I) 
DRVOL(I) 
MAXSTOR(I) 
MEANWTD(I) 



',7(F10.5 
\7(F10.5 
\7(F10.5 
',7(F10.5 
',7(F5.2 
•,F5.2) 
',7(F10.5 
: ',7(F10 
',7(F10.5 
',7(F10.5 
',7(F10.5 
',7(F10.5 
',7(F10.5 
',7(F10.5 



7A12) 
i 



X-NODE BOUNDARIES 
Y-NODE BOUNDARIES 
PARTIONING COEFF 
BULK DENSITY 

COMMENT : ',7A12) 



2X)) 
2X)) 
2X)) 
2X)) 



2X)) 
2X)) 
2X)) 
2X)) 
2X)) 
2X)) 
2X)) 

2X),/,2X,110('-')) 

2X)) 

2X)) 

2X)) 

2X) ,A8) 
5,2X) ,A8) 
2X) ,A8) 
2X) ,A8) 
2X) ,A8) 
2X) ,A8) 
2X) ,A8) 
2X) ,A8) 
,213) 
,713) 
,F7.5) 
,F7.3) 



190 



Figure C.l Continued. 
CCCCOC 

C THE END ! ! ! ! ! 
COCCCC 

END 

COCCCC 

C FUNCTION TO CALCULATE THE VOLUME DRAINED IN CM FOR A GIVEN WATER TABLE 
C DEPTH. FUNCTION CREATED BY USING LINEAR REGRESSION ON THE VOLUME 
C DRAINED - WATER TABLE DEPTH CURVE FOR WT DEPTH VALUES IN THE RANGE 
C OF 1.5 - 3.5 METERS 
COCCCC 

FUNCTION VOLDRAIN (WTDEPTH) 

VOLDRALN = -12.5887 + 29.048 * WTDEPTH 

RETURN 

END 







<-\ i 






























zw ■ 






























OA) • 




























































1— L. 1 






o 


fNJ 








f> 




in 


o 


in 






O CD" 


o 




f\J 


o 


ro 


ro 


ro 


f> 


in 


« — 


o 


o 






LU<1> I 




ro 


f\J 


fNJ 


fNJ 


fNJ 








, — 




fNJ 






QTD i 






























■ «W 1 






























O i 






























- 

DE 


(\i 


(M 




^ 


fNJ 


fNJ 


fNJ 


, 


* — 


fNJ 


fNJ 


(NJ 






i 3 • 










o 


O 


o 


o 


o 


O 


o 


o 






i 1- i 


















■ 


i 


1 


1 








UJ 


UJ 


UJ 


UJ 


in 


UJ 


UJ 


UJ 


UJ 


UJ 


LU 


UJ 






■ Z • 


ro 


^> 




CO 


in 




sO 


m 






in 








'-XJ i 




o 


o 






fNJ 


in 


fNJ 


m 


fNJ 


ro 


CO 






E< • 


%o 


ro 




o 


CO 




ro 


o 




f> 


in 








M i 


UN 


CO 


^ 




CO 


f> 


ro 






o 




CO 






E > 






























w 1 






























PE 








, 


OJ 


fNJ 


fNJ 


, 


fNJ 


fNJ 


fNJ 


fNJ 






o • 


o 


o 


o 


o 


o 


o 


O 


o 


o 


O 


O 


O 






_i i 
















1 


1 




i 


1 






CO 1 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 






>- 1 


IO 




ro 


ro 




CO 




fNJ 


f> 


in 


o 


-J- 






UJ i 




ro 


o 


o 


o 


ro 






o 


f> 


ro 


fNJ 






1 1 


r\J 




ro 


NO 




o 




o 


o 


fNJ 


-O 


O 






CO ■ 


*~ 




CO 








*~ 


*~ 




o 


in 


ro 




































t— 1 


( 




, 






, 












i 






QC • 






























UJ ■ 










fNJ 


fNJ 


fNJ 


fNJ 


, 


fNJ 


fNJ 


fNJ 






t— I 


o 


o 


o 


O 


o 


O 


O 


O 


o 


O 


O 


o 






<x ■ 


1 


■ 




1 


■ 


I 


1 


i 


1 






1 








UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 


UJ 








o 






O 


ro 


o 


ro 


o 


-J- 


CO 


CO 


CO 








o 


o 


ro 


in 




CO 


CO 


fNJ 


o 


N. 


o 


CO 








in 




(M 


fNJ 


CO 


o 


CO 


vO 


o 


CO 


ro 


CO 


< 




; ; 


in 


in 




fNJ 


-J- 


in 


fNJ 






fNJ 


ro 




J— 






























<£ 






, 






, 




, 








, 




, 


O 




• j 






























— ( 






























I ^ 1 


























CO 




Q_ £ , 




N. 


o 


o 




o 


ro 




m 


O 






<; 


00 




























i>— 


\ 


QV 1 




ro 






o 


fNJ 


>j- 


fNJ 


ro 


f> 


, 


>o 




in 




ro 


ro 










ro 


ro 


fNJ 




fNJ 




fy 


o 


|— • 


























LL> 


*s 




























»— 


in 




























< 




( ( 


N. 


r*- 


o 




fNJ 


o 




ro 


<J- 


CO 


fNJ 


f> 




ii 


• 1 • 


ro 


O 


o 


o 




ro 


ro 


ro 


fNJ 


« — 




CO 


UJ 




i 3 • 


o 


o 


o 


o 


, 


fNJ 


f> 


f> 


O 


fNJ 






o 


UJ 




























o 


t- 


| 


CO 




CO 


CO 


CO 


CO 






CO 


CO 


CO 


CO 




<^ 


E • 


r\j 


fNJ 


f\J 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


Q£ 




w i 


























oe: 






o 


O 


o 


in 


, — 


O 


ro 


o 


f> 


o 


O 


fNJ 


UJ 




ZCL i 






o 




LH 




f> 


o 


CO 


fNJ 


fNJ 


fNJ 






OZ • 


N- 




o 


f> 


f> 


*- 


m 


o 




o 


O 


O 


Q_ 

o 




1— I 


N- 




CO 


N- 


h- 


CO 


N- 


r — 




CO 


N- 


CO 


CD 




< I 


ru 


ru 


r\t 


ro 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 


fNJ 






> ■ 






























UJ i 

-JUL. ■ 


o 


in 


o 




o 


o 


CO 


o 


ro 


o 


o 








man i 


fNJ 




o 


ro 


r^. 




o 




CO 




fNJ 








■ Z> « 








ro 


OJ 


fNJ 


ro 


ro 


ro 




ro 


fNJ 




































■ ■ 






o 


o 


o 


o 


o 


o 


o 


o 


o 


o 






• i 


ro 


ro 


ro 


ro 


ro 


ro 


ro 


ro 


ro 


ro 


ro 


ro 








in 


in 


in 


m 


in 


in 


-* 






-J- 










E>- • 














ro 


ro 


ro 


ro 


ro 


ro 








































ro 


ro 


ro 


ro 


ro 


m 


in 


in 


in 


in 


in 




















CO 


CO 


CO 


co 


CO 


CO 






Z i 






























O • 




























































t— i 


o 


o 


CO 


CO 




O 


o 


O 


CO 


CO 


r>- 


NO 






<X i 


o 




ro 


in 




O 


o 




ro 


m 


N. 


f> 






u ■ 






























o ■ 




fNJ 


vj- 


^0 


CO 


o 




fNJ 


"■J" 




co 


o 






-J 1 






f\i 


ro 




SO 






fNJ 


ro 




NO 








o 


O 


o 


o 


o 


o 


o 


o 


O 


o 


o 


o 






o ' 


o 


O 


o 


o 


o 


o 


o 


o 


O 


o 


o 


o 




















CO 


CO 


CO 


CO 


CO 


CO 






z 






























o ■ 


i 


■ 


• 


i 






1 


















o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 






1 1- t 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 




































1 1— 1 


sO 


1^- 


CO 


o 


o 




•o 


r^ 


CO 


o 


o 








• CO 1 






























• ~a i 






























UJ i 


(\J 


OJ 


CM 


fNJ 


fNJ 


fNJ 


o 


o 


o 




o 


o 






o ■ 






























o— « 






























z 


ru 


o 


O 


-J 


CO 


fNJ 


fNJ 




o 




CO 


fNJ 


















fNJ 












fNJ 



191 





—1 




























1 




>> ■ 


























1 


o 


CO < 


-o 


>o 


o 




CO 


CM 


O 


m 


in 


CM 


«- 


CM 


1 


)— 1 


■o • 


o 


CM 


o 


ro 




CO 


-3- 


o 


ro 


CM 


r- 


in 


1 


2 




ro 


CM 


o 


o 




sj- 


sT 


m 


ro 


CM 


■o 


CM 


1 


UJ 




























I 




E i 


i 


' — 




«- 


i 


t 




ro 


CM 


«~ 




1 




CJ 


> — - i 




I 






















1 

z 


>- 


r~. | 


























o 


i— 


>- 1 




























i— 


CO 1 


O 


00 


in 


ro 


CO 


O 


N- 




■o 


in 


o 


»— 


i- 


o 


"D ■ 




ro 


CM 


CM 


co 


o 




o 




«* 




>* 


o 


o 


\ i 


o 


CM 


ro 


O 


CM 


ro 


O 


ro 


CM 


CM 






UJ 


-J 


E ■ 




























UJ 


w i 


1 




i 


i 


1 


■ 












1 


1 — ■ 


> 




























o 




^ i 




























LU 


>« i 


























>- 




CO ■ 


























1 


< 


T3 i 


00 


o 


o 


CM 


* — 


• — 


in 


« — 


* — 


ro 


CM 


-o 


1 


or 


i 


» — 


in 


o 


O 


in 




CM 




t-- 


in 


in 


CM 


1 




ro ■ 


o 


o 


o 


« — 


o 


o 


O 




o 


o 


o 


o 


1 


O 


« 


























1 


-J 


« ■ 


■ 


i 




1 


i 


i 












' 


1 


u_ 


E i 






























W 1 




























—J 






























< 


f-^ 1 


























1 


(J 


>> 1 




o 


o 


ro 


o 


o 


O 




>o 


o 


-J- 


CO 






CO i 


ro 


CM 


o 


r~- 


o 


o 


ro 






m 


o 




1 


Z 


~o • 


ro 


* — 


o 


CM 




ro 


r- 


in 


CM 


in 


ro 


in 


1 


UJ 




























1 


:c 


O) I 


i — 


i — 




i 


■ 


t 




i 


in 


i 




i 


1 


o 


E 


i 
















i 








1 




w 


























z 


>- 




























o 


1— 


1 


























1 — 1 




>« 


-j- 


CO 


T — 


CO 


co 


CO 


CM 


ro 


sf 


CM 


O 




>- 


<J 


CO 


«— 


T 


CO 


CO 


CO 


o 


i — 


•o 


CM 


* — 


CM 


o 


o 


O 


"D 


CM 


CM 


CM 


o 


1 — 


» — 


• — 


o 


O 


< — 


' — 


ro 


UJ 


_i 


^ 


• 


• 


• 


• 


• 


• 


■ 


• 


• 


■ 


■ 


• 




UJ 


E 


■ 




■ 


1 


1 


< 




• 




• 




• 


— • 


> 




























a 




























































X 


UJ 


>» 


























i 




CO 




-J" 


o 


-f 


ro 


•o 


CM 


ro 


» — 


-J" 


O 


-o 


i 


< 


T3 


r>- 


in 


o 


r— 


ro 


CM 




CM 




CM 


ro 


in 


• 


or 


v. 


o 


o 


o 


o 


o 


O 


o 


o 


' — 


O 


o 


o 


■ 


3 


ro 


























i 


O 


* 


■ 






1 


i 


• 






1 


■ 




i 




_j 


* 


























t 


u_ 


E 






























w 


































CM 


o 


o 


ro 


i — 


■ — 


* — 


■J- 


o 


o 


in 




—I 






o 


o 




ro 


o 


CM 




in 


o 


in 


CM 






























*o 


LU 


1- 


E 


























O 


O 


— ' 


ro 


o 






o 


in 


vl- 


ro 


r; 


CO 




-o 


^ 


*" 




C\J 












CM 


ro 








































O 




























































1/5 


_J 


*-\ 


-O 


o 


o 




1 — 


o 




*— 


o 


vO 




r^. 




i — i 


Ol 


o 


00 


o 




' — 


CM 


ro 


• — 


-o 


CO 


' — 






o 


E 




ro 


o 


O 


r» 


O 


ro 


o 


in 


CM 


o 


> — 


—J 


CO 




























< 






~* 


ro 




CM 


*~ 


' 


vj- 


■o 


u-i 


ro 


CM 
































































2" 


or 




























UJ 


UJ 






























>— 


o> 


o 


-o 


o 


-O 


CM 




ro 


o 


in 


in 


^r 


CO 


c_) 


< 


E 




r-. 


o 


O 


CM 


<1 


CO 


ro 


CO 


* — 


ro 








• — - 


o 


•o 


o 




O 


r- 


o 


CO 


r- 


CM 


ro 


vr 








o 


LTl 




o 


S- 


-j- 


o 


N- 


in 


in 


o 


m 








* — 


' — 










CM 


CM 


CM 


' — 




































< 






























o 


• 




co 




ro 


in 


CM 


o 


O 


r-- 






ro 






<-> 


A 


ro 


-o 


o 


o 


r— 


co 


in 




r^ 


o 


o 




z 


z 


a. 


























UJ 


o 


Q. 


N. 


o 


CM 


O 


^r 


' — 




in 


CM 


CM 


CM 


o 




o 


w 


w— 


CM 


CM 


* — 


r — 


i — 


» — 


CM 


ro 


CM 


• — 


































ED 


UJ 




t- 


in 


o 


o 


CO 




O 


o 


r- 


CM 


o 


r— 


i— 


C3 




ro 


o 


o 


r-- 


CO 


o 


r~- 


CM 


-o 




CM 


o- 


< 


< 


ro 


o 


in 


o 


o 


ro 


o 




O 


CO 


O 


CM 


in 




or 


* 


o 


N. 


o 




in 






o 


N. 


-O 


r-- 


in 




o 


* 


























*— 


}— 


E 


* — 












i — 


T 










< 


to 


























































o 






ro 


in 


ro 


CM 


,- 


ro 


o 


o 


in 




CM 


CO 


UJ 


UJ 






o 


r*- 


•O 


in 


in 


CM 


ro 


c^ 






CM 


z 


I 




•o 


ro 


o 


N- 


in 


in 


o 


CM 


O 




o 


CM 






ro 


CM 


ro 


ro 


C0 




CM 


ro 


vj- 




r- 






< 


_J 


« 


























or 


o 


* 


m 


in 


in 


-I 


•* 




in 


in 


in 




•<* 


<r 




> 


E 






























■ — • 


























I 


UJ 




<o 


CM 


O 


-O 


o 


O 


co 


O 


•O 


CO 


CM 


CO 


=> 


CJ 




ro 


in 


-O 


O 


•o 


ro 


o 


CM 


r^ 


ro 


in 


o 


Z 


< 




<o 




CM 


C0 


U1 


CM 


r>- 


O 


r- 


o 


o 


co 




or 


ro 


o 


CO 


o 


o 






CM 


CM 


-o 


o 


o 




X 


o 


* 


























< 


t— 


« 


o 


CO 


CO 




N- 


r-- 


o 


O 


CO 


r^ 


CO 






VI 


E 






























































<M 


CM 


CM 


CM 


CM 


CM 


o 


•O 




o 


v O 


-o 








CM 


■& 


O 




CO 


CM 


CM 


-O 


o 


•4 


CO 


CM 


















CM 








«— 


T— 


CM 



I 



OOOOOOOOOS-OOOOOOi 

ooooN-N-oomooooooo 

(OOOONNinOSCONOOOOOUM 
OOOO*0^N-O*©^)OOOOOv- 

coooomm*— coo«— ooooo 

















(NJ 


o 


m 






ini 












INI 


ro 


o 




CM 






CO 


r- 










I 


st 


t> 










(NJI 


-o 












sO 


ro 










>o 

1 


o 


o 


o 


ro 


o 


r- 


o 


o 


tNI 


r- o 


o o 


o o 


Oi 


fNi 


o 


o 


fNJ 


r-~ 


st 




•* 


O 


«- o 


o o 


o o 


o 


■o 


o 


o 


NT 


<\J 


CO 


o 


o 


ro 


r~- o 


o o 


o o 


Oi 


^* 


o 


o 


ro 


<o 


r\j 


ro 


in 


cr- 


<o o 


o o 


o o 


o 


in 


o 


o 


in 


in 


o 




r- 


co 


r- o 


o o 


O O Oi 


K 






<f 




N. 




sr 


r\i 


>» 
















r\i 


1 


sf 


o 


r- 


fNJ 








CO 












o 


o 










1 


o 


o 


o 


r- 


ro 




o 


o 




co o 


o o 


o o 


Oi 


<\j 


o 


o 


t\j 


r\i 


o 


•o 


o 




~J o 


o o 


o o 


o 


^o 


o 


o 


ro 




o 




o 


-j- 


■O o 


o o 


o o 


Oi 


<f 


o 


o 




ro 


o 


r- 


r\j 


m 


CO o 


o o 


o o 


o 


in 


o 


o 




in 


(NJ 




cr- 


o 


sJ- o 


o o 


o o 


Oi 


S. 






t> 


-J- 


in 


r- 


CO 


t-~ 


<o 






i 
















r- 




r\J 








CO 














m 










i 



ooOT-r^mooocoooinotn(\ii 

OOOOWKlO'O'flNOONON'- 

o*oo«-ro«-rooo«cooor\joc\ir--i 
T-ooosj-sj-oooooomOLn*o 

MOOOSSO»-vtOOONONOi 



*- «- ■ O CO «- <NJ 
(NJ N- *— 



r~- co i 



*-ooo«-(nj(\J(nj(nj 
ooooooooo 
■ + + + + + + + + 

i in in i i in 

tNir-co -jo-osoMomo 
^tT-rMoN-r^coco-j-cococoo 
■ ■ ■ ■ mo«-rN-om<000 
oeoeoinm*-»-Sfvji-Kiooo 

lONfXIS 



i-i-Ni-OIMi-OO 

ooooooooo 
I • ■+ + + + + + 
r\ir\jr\i ujujujujujujujujuj 
i-MNO>ono>eOint\jinoo 

OOWO-OONOO 

<o<oocM<OfMrooo 

Si-f-t-i-MCJOO 



N Oi- 
(NJ CM (NJ 





































z 


z 


z 




z 


z 




z 


z 


at 


a: 


LU 






































o 


o 


o 




O 


o 




O O 






O 


o 






















































a 


Q 


< 




o 


o 


o 






in 


o 


o 


O 


N- 


o 


o 


CO 


in 


ro 


Oi 




1— 


I— 


r- 




t— 


t— 




i— 


1— 


■ 




Ct 


r— 


o 


o 


o 


o 


i> 


(NJ 


o 


vt 


CO 


N. 


o 


o 




r-~ 


-J- 


(Nl 




< 


< 


< 




o 


u 




(_) 


f— > 


X 


>- 


o 


< 


o 


o 


o 








(NJ 


in 


in 


f> 


o 


o 


o 


fNJ 


<i 


•Oi 




> 


> 


> 


m 


LU 


LU 




LU 


LU 








at 




o 


o 


r-- 


O 


(NJ 




O 


f> 




o 


o 


m 


in 


o 


fNJ 




LU 


UJ 


UJ 


r— 


QC 


ac 


LU 


ad 


at 






CO 


}— 


ro 


o 


o 




o 


CO 


(NJ 


ro 


r- 


m 


o 


o 


<o 


fNJ 


st 


sf 1 




—1 


_l 


i 


a. 






o 












z 




































UJ 


LU 


UJ 


LU 


Q 


Q 


=> 


Q 


o 








LU 


o 






st 


s- 


(NJ 


-J- 


CO 


fNJ 


in 






o 




ro 


o< 










Q 


I 


i 


1— 


1 


1 


z 


Z 


CO 


CJ 


(NJ 










i — 


IN- 


(NJ 




fNJ 






ro 




(NJ 


■* 




UJ 








X 


>- 




X 


>- 






CO 


z 


(NJ 










1 


O 


O 
















•J-i 




o 




LU 


LU 






z 










< 


o 




































< 


Q£ 


—I 


— 1 






C3 










z 


CJ 


































I DC 


Li_ 


UJ 


CO 


CO 


z 


Z 


< 






3 


3 






































■ LU 


at 


>- 


< 


< 






z 


z 


z 


O 


O 


_l 


— 1 


o 


o 


o 


CO 




CO 


o 


o 


o 




o 


o 


o 


CO 


fNJ 


Oi 


I h- 




< 


1— 


1— 












— 1 


—J 


< 


< 


o 


o 


o 




-o 


vf 


o 


r^ 


~t 




o 


o 


N. 




in 


O 


■ UJ 


to 


















Li- 


Li- 


o 


O 


t> 


o 


o 


o 


r-- 


CO 


-o 


o 


o 


IN- 


o 


o 


CO 


o 


o 


O. 


i X 






an 




LU 


LU 


LU 
















o 


o 


r-. 




o 


in 


o> 


o 


r^ 


o 


o 


«— 


in 


in 


O 


i < 


—J 




LU 


LU 


O. 


O- 


a. 


3 


3 


z 


Z 


z 


z 


ro 


o 


o 


r~- 




ro 


-o 


(NJ 


ro 


o 


o 


o 


CO 


-o 




tN-l 


I OC 




o. 


r— 




O 


O 


o 


o 


o 


LU 


LU 


LU 


LU 




































o 


z 


< 


< 


_J 


_l 


—i 


1 


_l 


zr. 


3= 


3C 


31 


-o 










(NJ 


O 


CO 


co 


O 






o- 


o 


f> 


Oi 


> a. 


CO 




3 


3 


CO 


CO 


to 


u- 


u_ 


O 


(-> 


CJ 


(J 


(M 










• 


CO 


(NJ 


fNJ 


(NJ 






•o 


ro 


ro 


o 






























tNI 












o 


o 
















(NJl 




























































ro 

i 






























O 


o 


o 


o 


CO 


CO 


o 


o 


in 


o 


o 


o 


o 


o 


o 


Oi 






























O 


o 


o 


o 






o 


r^ 


ro 


r- 


o 


o 


o 


r~ 




o 






























in 


o 


o 


o 


t> 


o 


ro 


CO 


in 


o 


o 


o 


o 


CO 


co 


■*! 






























o 


o 


o 


o 


r-~ 




O 


r- 


in 


o 


o 


o 


o 






fNJ 






























r- 


o 


o 


o 


r~- 


r~ 


o 


ro 


vl- 


ro 


o 


o 


o 


co 


CO 


fNJi 






























o 












(NJ 


fNJ O 


o 








o o 


Oi 






























ro 












-J- 


CO 


O 










-o -o 


st 












































o 














1 


mi 




























































(NJ 
















































*-\ 












i 
































- — ' 








" — * 


w 


■ — . 




' — * 




INJ 




(NJ 


3 


^i 






























. — V 




fNJ 




(NJ 


3 


Ot 








X 


X 


>- 


>- 


O 


wi 
































X 


X 


>- 


>- 


O 


O 




f/5 


r— 


3 


3 


3 


3 


—I 


CO 






























1 


3 


3 


3 


3 


_l 


1— 


_l 


in 


3 


o 


o 


o 


o 


u_ 


t/>l 
































o 


o 


O 


O 


u- 


CO 


o 


< 


z 


_l 


_l 


1 


—i 


(J 


<c 






























UJ 


— 1 


_J 


_J 


_J 


t— 


X 


> 


z 














Zi 






























at 


u_ 


li- 


u_ 


u- 


UJ 


< 


at 


3 


UJ 


(_> 


CJ 


CJ 


CJ 


UJ 


CJ 






























< 


CO 


ra 


CD 


CO 


z 


z 


Q 


1— 


z: 


CO 


CO 


CO 


CO 


z 


t-l 































CO 

UJ 

-1 
o 



CO 

o 



T3 

0) 

C 



o 

o 



CO 

o 



3 

in 



o 
o 

«* 

> — 

o 
o 

S* 



o 
o 



o 
o 

s* 



o 

>»o 

CO 

"O o 
~s, cc 
E UJ 
N 

St 

T- O 

to Ul 

II CJ 
O Of 
Kl >- O 

I- IJ. 

•o -- 

<\i > o 

*s, 

O I- 3 

(M O 

3 < 

st Q H- 

•- z < 

o o o o 
O O IO CO o 
co in (M ui 

st . ■ Z 

O »-«-«- Q 

>- IM 





o 


1— 


o 




o 


O st 












< 






0£ 


< 






o: 


oe «- 












a: 






C3 


Q 


C3 




CJ 


CJ 










NT 


AT 






< 




< 




> 


MOO 






CO 


CO 


< 




o 




m 




UJ 




_j 


cc o m 


Ll_ 




UJ 


UJ 


i— 


cj 


in 




i 


UJ 


CJ 




UJ 


I— • • 


LU 








CO 


z 






Q£ 


-J 


< 




in 


< -* 


UJ 






a; 


z 




tsl 




UJ 




Li- 




o m 


O 




< 


< 


o 


CO 


r\i 




0. 


ll. 


Cf 






o 


CJ 


>- 


o 


Q 


CJ 


=> 




CO 


z 


■ 


=> 


CO 


m 






1— 


z 


Z 






<\j 


LLI 






CO 


UJ 




in 


CJ 




=> 








in 


_J 








-J 






Z 


CO 


o 


o 










II 


II 


II 




H 


II o >■ 




z 


m 


CO 








U_ 








u- 




Z 1— 


z 


UJ 






1— 


1— 


<\i 






(M 


Kl 




s* 


sO o — 


o 


o 


UJ 


LU 


z 


z 




1— 








t- 




CJ CO 






o 


o 


LU 


UJ 


i— 


=) 


1— 


1— 


1— 


=3 


i— 


1- o 


1— 




o 


o 


X 


s: 


1 


0- 








o. 




-out 


or 


1 


z 


z 


z: 


X 




Z 


z 


Z 


z 


z 


z 


z >- O 


< 


=) 


1 


1 


o 


o 






3 


3 


=> 




=> 


3 Z O. 


O- 


CO 


X 


>- 


CJ 


CJ CO 



00 

o 



sO 

o 
s 

SO 



o 

-Q 
CO 
H 



< 

CJ 
O 



194 



Figure C.2 Program to calculate mass balances between sampling periods. 
CCCCCC 

C PROGRAM TO ANALYZE MASS BALANCES BETWEEN SUCCESSIVE SAMPLING PERIODS 
C CREATED: 1/3/88 BY: MATT C. SMITH 
C LAST UPDATE: 4/6/88 
CCCCCC 

REAL NETWFLOW(28,7) ,NETCFLOW(28,7) ,TWMASS(28,7) ,TCMASS(28,7) 
REAL CERROR(28,7) ,WERROR(28,7) ,AVGWFLOW(28,7) ,AVGCFLOW(28,7) 
REAL DELWMASS(28,7) ,DELCMASS(28,7) ,AREA(7) ,PERC(28) 
REAL PDELWMASS(28,7) ,PDELCMASS(28,7) ,DRVOLSUM(28,7) ,MSTORSUM(28,7) 
REAL MEANWTD(28,7) ,DELTAWTD(28,7) ,WERRORP(28,7) ,CERRORP(28,7) 
REAL TOTWERROR(7) ,TOTCERRDR(7) ,CFLUX(28) ,APPAREA 
INTEGER I,J,DAY(28) ,NFILE 
CHARACTER*8 DATE (28) 

CHARACTER* 12 INFILE,INnLE2 / CiUTFII^,CX»1MENrS(100) 
CCCCCC 

C APPAREA = APPLICATION AREA SIZE (ha) 
CCCCCC 

APPAREA = 0.02787 
CCCCCC 

C READ IN INPUT AND OUTPUT FILENAMES 
CCCCCC 

OPEN (UNIT=1 , FILE= • INFLUX . DAT STATUS= 1 OLD • ) 

READ(1,102)INFILE 

READ ( 1 , 108 ) NFILE 

READ ( 1 , 102 ) INFILE2 

READ ( 1 , 102 ) OUTFILE 
102 FORMAT (A12) 
108 FORMAT (12) 
CCCCCC 

C OPEN FILES 
CCCCCC 

OPEN (UNTT=3 , FILE=INFILE , STATUS= 1 OLD 1 ) 
OPEN (UNTT=4 , FIL£=INFIIE2 , STATUS= ' OLD 1 ) 
OPEN (UNIT=5 , FILE=OUTFILE , STATUS= ' UNKNOWN ' ) 
OPEN (UNTT=7 , FILE= 1 SUMMARY . SUM ', STATUS= 1 UNKNOWN') 
CCCCCC 

C READ IN DATA FROM INPUT FILES 

C DAY (I) = DAYS SINCE APPLICATION 

C PERC(I) = PERCOLATION DURING PERIOD (cm) 

C CFLUX(I) = CHEMICAL FLUX DURING PERIOD (g/ha) 

C AREA(I) = AREA OF SUBAREAS (m**2) 

C NETFLOW(I) = NET WATER FLUX INTO (+) OR OUT OF (-) SUBAREAS (m**3/day) 

C TWMASS(I) = TOTAL WATER IN SUBAREA (m**3) 

C DRVOLSUM(I) = DRAINED VOLUME LN SUBAREA (cm) 

C MSTORSUM(I) = WATER IN SATURATED ZONE IN SUBAREA (m**3) 

C MEANWTD(I) = MEAN WATER TABLE DEPTH LN SUBAREA (CM) 

CCCCCC 

READ(3,721) (COMMENTS(I) ,1=1,54) 

DO 10, I = 1, NFILE 

READ(3,*) 



195 



Figure C.2 Continued. 

READ(3,115) (NETWFLOW(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (TWMASS(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (DRVOLSUM(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (MSTORSUM(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (MEANWTD(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (NETCFLOW(I,J) ,J=1,7) ,DATE(I) 

READ(3,115) (TCMASS(I,J) ,J=1,7) ,DATE(I) 
10 OCNTINUE 
CCCCCC 

C BEGIN CALCXJIATIONS 
CCCCCC 

WRTEE(*,94) 
94 FORMAT ( 1 BBGINING CAICUIATIONS 1 ) 
DO 15, I = 2,NFII£ 
DO 15, J = 1,7 

CCCCCC 

C CAIUIATE AVERAGE WATER AND CHEMICAL FLUX IN EACH SUB-AREA BETWEEN 

C SAMPLING DATES 

CCCCCCC 

AVGWFLOW(I,J) = (NETWFLOW(I-l,J) + NETWFLOW(I, J) ) / 2.0 
AVGCFLOW(I,J) = (NETCFLOW(I-l,J) + NETCFLOW(I, J) ) / 2.0 

CCCCCC 

C CALCULATE CHANGES IN OBSERVED PARAMETERS BETWEEN DATES 
CCCCCC 

C DRAINED VOL § I GT DRVOL § 1-1 => WATER LOSS 

DELWMASS(I,J) = DRVOISUM(I-l,J) - DRVOLSUM(I, J) 
DELCMASS(I,J) = TCMASS(I,J) - TCMASS(I-1,J) 
DELTAWTD(I, J) = MEANWTD(I, J) - MEANWTD(I-1, J) 

CCCCCC 

C CALULATE PREDICTED CHANGE IN WATER AND CHEMICAL MASS BASED 

C ON AVERAGE FLOWRATES BETWEEN PERIODS 

CCCCCC 

PDELWMASS(I,J) = AVGWFIOW(I,J) * FLOAT (DAY (I) -DAY (1-1) ) + 

# (PERC(I)/100.0)*AREA(J) 
IF(J.EQ.2) THEN 

PDELCMASS(I,J) = AVGCFLOW(I,J) * FLOAT (DAY (I) -DAY (1-1) ) + 

# CFLUX(I) * APPAREA * 1000.0 
ELSE 

PDELCMASS(I,J) = AVGCFLOW(I,J) * FLOAT (DAY (I) -DAY (1-1) ) 
ENDIF 

711 FORMAT (A8, 413, 8 (F8. 3, IX)) 
CCCCCC 

C CALCuTATE DIFFERENCE BETWEEN PREDICTED AND OBSERVED VALUES 
CCCCCC 

WERROR(I,J) = PDELWMASS(I,J) - DELWMASS(I, J) 
CERROR(I,J) = PDELCMASS(I,J) - DELCMASS(I,J) 
TOTWERROR(J) = TOTWERROR(J) + WERROR(I,J) 
TOTCERRDR(J) = TOTCERROR(J) + CERROR(I,J) 



Figure C.2 Continued. 



CCCCCC 

C CALCULATE THE PERCENTAGE ERROR FROM OBSERVED VALUE 
CCCCCC 

IF(DELWMASS(I,J) .NE.O.O) THEN 

WERRORP(I,J) = 100.0 * WERROR(I,J) / ABS(DELWMASS(I, J) ) 



ENDIF 

IF(DELCMASS(L,J) .NE. 0, 
CERRORP(I,J) = 100.0 
ENDIF 
CONTINUE 



0) THEN 

* CERRDR(I,J) / ABS(DELCMASS(I,J)) 



15 
CCCCCC 

C WRITE RESULTS TO OUTPUT FILES 

CCCCCC 

WRirE(5,125) 

WRITE(5,721) (COMMENTS(I) ,1=1,54) 
WRTTE(7,721) (COMMENTS (I) ,1=1,54) 

DO 25,I=1,NFILE 

IF(I.EQ.l) THEN 

C FOR 1ST DATA PERIOD THE ARE NO COMPARISONS TO BE PRINTED 

, 130) DATE (I) , ' 11/12/86 ' 
, 135) DAY (I) , 1 11/12/86 ', PERC (I) 
,137) (AREA (J) ,J=1,7) 
f 241)DATE(I) , (MSTORSUM(I,J) ,J=1,7) 
,242) DATE (I) , (DRVOLSUM(I, J) ,J=1,7) 
,145)DATE(I) , (TWMASS(I,J) ,J=1,7) 
,243)DATE(I) , (MEANWTD(I, J) ,J=1,7) 
,140)DATE(I) , (NETWFLOW(I,J) ,J=1,7) 
,136) DAY (I) , 1 11/12/86 •,CFLUX (I) 
,150)DATE(I) , (NETCFLOW(I / J) ,J=1,7) 
,155)DATE(I) , (TCMASS(I,J) ,J=1,7) 

130) DATE (I) ,DATE(I-1) 

135) DAY(I) ,DATE(I-1) ,PERC(I) 

241) DATE (I) , (MSTORSUM(I,J) ,J=1,7) 

242) DATE(I) , (DRVOLSUM(I, J) ,J=1,7) 
145)DATE(I) , (TWMASS(I,J) ,J=1,7) 

243) DATE(I) , (MEANWTD(I, J) ,J=1,7) 

140) DATE(I) , (NETWFLOW(I,J) ,J=1,7) 

141) (AVGWFLOW(I,J) ,J=1,7) 
247) (DELTAWrD(I,J) ,J=1,7) 

142) (DELWMASS(I,J) ,J=1,7) 

143) (PDELWMASS(I,J) ,J=1,7) 

144) (WERROR(I,J) ,J=1,7) 
177) (WERRORP(I,J) ,J=1,7) 

136) DAY (I) ,DATE(I-1) ,CFLUX(I) 

150) DATE(I) , (NETCFLOW(I,J) ,J=1,7) 
155)DATE(I) , (TCMASS(I,J) ,J=1,7) 

151) (AVGCFLOW(I,J) ,J=1,7) 

152) (DELCMASS(I,J) ,J=1,7) 

153) (PDELCMASS(I,J) ,J=1,7) 



WRITE ( 


5, 


WRITE i 


5 


WRITE) 


5 


WRITE 


5 


WRITE 


5 


WRITE! 


5 


WRITE 1 


5 


WRITE! 


'5 


WRITE! 


'5 


WRITE 


'5 


WRITE 


:5 


ELSE 




WRITE 


[5 


WRITE 


[5 


WRITE 


15 


WRITE 


[5 


WRITE 


[5 


WRITE 


(5 


WRITE 


[5 


WRITE 


[5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 


WRITE 


(5 



Figure C.2 Continued. 



WRITER, 154) (CERRDR(I,J) ,J=1,7) 
WRnE(5,178) (CERRORP(I,J) , J=l,7) 
WRTTE(7,135)DAY(I) ,DATE(I-1) ,PERC(I) 
WRITE (7, 141) (AVGWFLOW(I,J) ,J=1,7) 
WRTrE(7,142) (DELWMASS(I,J) ,J=1,7) 
WKnE(7,143) (PDELWMASS(I,J) ,J=1,7) 
ENDIF 
25 CONTINUE 

WRITE(5,436) (TOTWERROR(J) ,J=1,7) 
WRITE (5, 437) (TOTCERROR(J) ,J=1,7) 
WRITE (7, 438) (TOTWERRDR(J) ,J=1,7) 

CCCCCC 

C STOP AND END PROGRAM 
CCCCCC 

STOP ' END OF PROCESSING !!!!!' 
CCCCCC 

C FORMAT STATEMENTS 
CCCCCC 



16X,7(F10.5,2X) ,A8) 
(• SUMMARY OF FLUX DATA 1 ,/) 
f , • CHANGE BETWEEN 1 ,A8, 1 AND 1 ,A8) 

DAYS SINCE APPLICATION: ' ,13, ' , PERCOLATION SINCE 1 , 
'^5.2,' cm 1 ) 

DAYS SINCE APPLICATION: 1 ,13, 1 , CHEMICAL FLUX SINCE 
',F5.2,' mg/m^') 

: ',8(F10.4,2X)) 
\8(F10.4,2X)) 
',8(F10.4,2X)) 



110 


FORMAT (! 


115 


FORMAT (: 


125 


FORMAT 


130 


FORMAT (, 


135 


FORMAT ( 




# A8," = 


136 


FORMAT ( 




# AS, 1 = 


137 


FORMAT ( 


140 


FORMAT ( 


141 


FORMAT ( 


142 


FORMAT ( 


247 


FORMAT ( 


143 


FORMAT ( 


144 


FORMAT ( 


145 


FORMAT ( 


242 


FORMAT ( 


241 


FORMAT ( 


243 


FORMAT ( 


155 


FORMAT ( 


150 


FORMAT ( 


151 


FORMAT ( 


152 


FORMAT ( 


153 


FORMAT ( 


154 


FORMAT ( 


177 


FORMAT ( 


178 


FORMAT ( 


436 


FORMAT ( 


437 


FORMAT ( 


438 


FORMAT ( 


721 


FORMAT ( 




END 



AREA OF SUEAREA (m 2) 
NETWFLOW ON \A8, 1 : 
AVGWFLOW OVER PERIOD 
DELTA WSTORAGE (irT3) 
DELTA MEAN WTD (cm) 
PRED DELTA WSTORAGE 
ERROR IN WSTORAGE (m' 3) 



SAT. STORAGE ' ,A8, 1 
DRAINED VOL 1 ,A8, 1 
MAX STORAGE ' ,A8, 1 
MEAN WTDEPTH 1 ,A8, 1 
C STORAGE ON 1 ,A8, 1 
NETCFLOW ON 1 ,A8, ' 
AVGCFLOW OVER PERIOD 
DELTA CSTORAGE (mg) 
PRED DELTA CSTORAGE 
ERROR IN CSTORAGE 
ERROR IN WSTORAGE (%) 
ERROR IN CSTORAGE (%) 
/, 1 TOT ERROR IN WSTORAGE 
TOT ERROR IN CSTORAGE : 
TOT ERROR IN WSTORAGE : 
6A12) 



',8(F10.4,2X)) 
•,8(F10.4,2X)) 
',8(F10.4,2X)) 
\8(F10.4,2X)) 
»,8(F10.4,2X)) 
',8(F10.4,2X)) 
',8(F10.4,2X)) 
',8(F10.4,2X)) 
'^(FIO^^X)) 
• ,8(F10.4,2X)) 
',8(F10.4,2X)) 
\8(F10.4,2X)) 
•,8(F10.4,2X)) 
■ f 8(F10.4 # 2X)) 
',8(F10.4 f 2X)) 
• f 8(F10.4,2X)) 
: ',8(F10.4,2X)) 
•,8(F10.4,2X)) 
',8(F10.4,2X),/) 



198 



o co 
•- t- 
s* ro 



o co 

<0 (M 



o «- 

ro «- 

o <o 

<m »- 

rs. «- 

• CO 



ro -O 

co -O 

<0 o 

ro «- 

st r«- 



o o 

(M r«- 

ro «m 

in ro 

co <o 

i r«- 



<m co 

in st 

«- o 

ro o 

ro m 



o o 

co m 

m o- 

ro co 

in in 



o> oj 

ro & 

•© oj 

•O o 

st o 



■o o 

CO r- 

-O -O 

in ro 

m >0 



in ro 

«- -O 

s* ro 

in o 

in «- 



«- O 

o O 

<\j st 

in t— 

oj 

ro 



o «- 

ro ro 

co st 

-o o 

ro in 



■O in 

o o 

m co 

s» O 

ro ro 



o- r- 

o- sf 

m m 

<m ro 

m cm 



o o 

•o o 

o o 

O m 

ro st 



m 

ro st 

st oj 

>* o 

r\j o 



X 
P 
►J 
Pm 

s 

M 

60 

o 
rJ. 

ft 

§ 

M 

M-l 

4J 

3 

O 

a) 
i — i 
a 



co 

CN 

U 

QJ 
i — I 

id 

H 



o 

OJ 



EA 






o 
o 








ac 


< 












< 


1— 




ro 








CO 


< 


o 








o 


z> 




in 








ro 


co 
















X 


o 


o 










=) 


ro 


o 






ro 




—J 














u_ 


ro 


ro 






ro 


>- 




st 








ro 


1— 


z 














o 


O 








CO 


> 






o 














o 








1— 


< 


1— 








st 


CJ 


_J 


< 


ro 








Z> 


o 












Q 


CJ 








CO 


to 


z 


oc. 


co 










o 


LU 


CO 


o 








u 


a. 


o 


o 




o 


ro 


13 


z: 


V, 


ro 








z 


M 


OJ 












oc 


ro 










M 


a. 


*s 














st 


o 


Sj 


00 


CO 


3 


03 


o 


o 


ro 


LU 


LU 




z 












1— 




z 


ro 




OC 


0C 


o- 


CO 


o 






< 


< 


O 


=) 


RUN 






OUND 


OUNO 








Q 


>- 


co 


CO 


1— 






z 


i— 








z 


< 


O 




LU 


LU 


LU 


UJ 


oc 


oj 


00 


o 


o 


z 




03 




O 


O 


o 


I 


s 


o 


Q 


OC 


z 


z 


o 


o 


a: 


>- 


o 


t 


1 


u 


u 


a. 


=c 


a. 


X 


>- 



ro ro 

«- o 

o -o 

o o 

E in ^ 

O «- <M 
CM 

ro 



ro «- 

m >0 

II CO nT 

«- <o 

o • • 

00 O h- 

"v *- I 

m i 



LU O O 

CJ in o 
z m ro 



st ro 
o o 
ro o 



ro 
ro 



st 
o 



ro 



co o 

CO in 

ro co 

o. o 

-o • • 
co «- 



ro 
ro 



ro 



ro 



rs. 

LU <0 
CJ O OJ 

z st rs. 



ro 



< 

-i ro O 
O «J *J 
cj in vi- 
ae ro ro 
uj • ■ 
a. ro ro 



o 

st 



o 

r— ro uj 

< < o 
u E < 
*~ w ac 
-I o 

D_ LU I— 

a. o co 

< < 3 

at 

LU O < 

CJ I — I — 

Z CO — I 

— 3 UJ 

CO Q 
< 

V) h O 

>— —J UJ 

< UJ oc 

a a a. 



CO 



< 

_l O O 

o oj co 

u <t i- 

cc -a co 



rs. 
st 



i- ro uj 

< < u 
CJ E < 
_ w or 
-I O 

d 111 H 

a. cd oo 

< < 3 
OC 

LU O < 

CJ r— r— 

Z CO — I 

3 uj 

CO o 

< 

WHO 

>■ J u 

< UJ oc 
Q Q O. 







O 




ro 


St 




O 


ro 




ro 


O 




o 


in 




<o 


O 




o 


ro 




• — 


ro 




ro 


O 




ro 


CO 






st 




•o 


in 


E 


in 


o 


E 




rs. 


E 


ro 


st 


u 


i 


ro 


o 




ro 


o 




st 


ro 






in 






00 






O 






ts. 






ro 






ro 






ro 






in 








o 






in 


rs. 




o 


ro 




o- 


ro 




ro 


ro 




CO 


CO 


ii 


rs. 


o 


ii 


vO 


ro 


n 




CO 




nO 


ro 




St 


o 




■o 


ro 


-O 






rs. 






rs. 






CO 


•o 


ro 


00 




■o 


00 


st 


ro 








s 












o- 






in 






ro 






ro 






o 


















V 












ro 






































ro 


ro 






ro 




in 


O 


UJ 


in 




LU 


st 


ro 


UJ 


o 




CJ 


o 


o 


CJ 


rs. 


rs. 


OJ 


CO 


o 


z 


o 


o 


z 




in 




in 




00 


ro 


ro 


00 




ro 


00 


in 




















ro 


z 


< 




z 






z 






o 






o 






o 
























1— 






1— 






l— 






< 






< 






< 






—1 


co 


st 


—1 


in 




_l 


00 


oo 


o 


o 


OJ 


o 


o 


rs. 


o 


is. 


CO 


CJ 






CJ 


o 


ro 


OJ 


st 




ac 






oc 




CO 


oc 


ro 


oo 


UJ 






LU 






LU 






O- 


00 


o 


a. 




ro 


o- 


st 


CO 




1 














ro 




















st 












co 






in 






>o 






•o 






z 






z 






z 






o 






o 






o 






1— 


ro 


LU 


1— 


ro 


LU 


1— 


ro 


LU 


< 




03 


< 




03 


< 




03 


CJ 


E 


< 


CJ 


'e 


< 


CJ 


■ 


< 




\*r 






> — ' 


QC 






QC 


1 




O 


i 




o 


— 1 




o 


O- 


LU 


r— 


a. 


LU 


1— 


0- 


LU 


1— 


a. 


03 


00 


a. 


03 


00 


a. 


03 


CO 


< 


< 


3 


< 


< 


3 


«t 


< 


3 




OC 






QC 






QC 




LU 


o 


< 


LU 


O 


< 


LU 


O 


< 


CJ 


1- 


l— 


CJ 


1— 


1— 


CJ 


I— 


I— 


z 


to 


_l 


z 


CO 


—I 


z 


oo 


— I 




3 


LU 




3 


LU 




3 


LU 


UJ 




Q 


00 




Q 


CO 




Q 




< 






< 






< 




CO 


1— 


O 


00 


t— 


O 


00 


1— 


Q 


>- 


_l 


LU 


>- 


_l 


LU 


>- 


—J 


LU 


< 


LU 


OC 


< 


LU 


QC 


< 


LU 


QC 




Q 


B, 


a 


Q 


O. 


Q 


O 


a. 



199 





O* 














in 


cm 




CO 








is. 


o 




so 


o 








^ 






1 






CM 


o 




i 

st 


1 






O 


in 












•3 






is. 








so 








CV 






ro 














so 


♦ 






[ 


































o 






o 


CO 






co 


CO 




o 


o 






o 


CO 






CO 






cm 


-o 




o 


ro 






so 


r- 




c> 






















00 


rs. 




in 


ro 








in 




o 


o 






o 


rs- 




-J 


ro 






CM 


-o 




CM 


o 




E 


CM 


ro 


e 




CO 


e 


o 






o 






o 








o 






«* 


st 






o 






in 














?s! 




ro 


vO 




O 


t\j 






o 


CO 




O 


is. 




II 


o 


<SI 


II 


O 


CM 


ii 




CO 


«* 




CM 


o 




1s. 






N. 






r-~ 


CO 


S-' 


INI 


CO 


■ 


CM 


CO 


*s 




1 


"V 




1 — 


ss 


ro 






LO 




1 


CO 


o 






o 






o 














"s. 


LO 






in 






in 



o 


ro 




rs. 




o 


CO 


o 


CO 


o 


ro 


o 


o 


sf 




CO 


ro 


m 


rs. 


ro 


ro 


ro 


N- 


in 




o 


o 


CM 


in 




o 


in 


o 


^r 




rs. 


O 


ro 


st 


in 


CO 


o 


o 


SO 






o 


r-! 


o 


ro 




ro 


so 










st 




rs. 




CM 


1 










i 


1 


rs. 












■ 










rs. 


CO 


, — 


, — 




o 


o 


ro 


CM 


-o 


CO 


CM 


o 


O 




in 


o 


CM 


O 


in 




vO 


co 


o 




CO 




so 


o 


O 


r-- 


O 


CO 


ro 




so 


r^ 
• 


o 
• 


in 
• 


st 

• 


o 






ro 




in 


O 


st 


ro 


CO 


• 


in 




CM 






CM 






o 










i 




i 




in 


rs. 




r- 


CO 




CM 


O 


CM 




rs. 


o 


st 


o 


m 




O 


o 


ro 


so 


CO 


st 


CO 

ro 


CM 


Cr- 




O 




o 


in 


in 


•G 


CO 


st 




in 


co 


o 


SO 


o 


rs! 


co 








st 


CM 


o 


CM 


rs! 




CM 




1 




CM 
1 


CM 
1 




CM 
i 


o 
i 


O 


in 


is. 


O 




CM 




co 


o 


in 


-O 


o 


CO 


CO 




CO 


o 


o 


in 


so 


rs. 


SO 


CO 


O 




r- 


r- 


o 


N. 


ro 


O 


o 




CO 




in 


O 


st 


CM 


ro 




o 


E C* 


CM 


£ 


CM 


ro 


E ro 


st 


o 




in 


U I 




o 


CM 




u ■ 




CM 










1 








so 






st 




ro 






o 










S3 




r- 






o 








CM 


o 


CO 




CO 


O- 


st 


sO 


o 


in 


CM 


<o 


O 




o 


st 


CM 


o 


st 




N. 


II CM 


ro 


ii 


rs. 


CO 


II to 


00 


st 


sD 


O 


O 


|s- 




st 


CO 


O 


CO 


S3 






r- 




r- 






r- . 






NT 


(O 


CO CM 




co 


in 


sO 


CO st 


st 


n! 






>s 








1 


\ i 









SO 


in 




O 


o 




|s- 


rs. 










CM 


CM 




st 


CO 


Is. 


UJ 


st 


st 


UJ 


o 


st 


LU 


st 


ro 


UJ 


CM 




UJ 


CM 


st 


UJ 


in 


o 


CO 


CJ 


st 


CM 


cj 


S3 


o 


CJ 






u 


to 


CM 


CJ 


rs. 


ts. 


CJ 


ro 


o- 


O 


z 




st 


z 


in 


o 


z 




o 


z 


CM 




z 


oo 






ro 


st 


C0 


00 


O 


o 


00 


ro 


CM 


00 


so 


•o 


oo 


O 




00 


ro 




00 


o 


in 


CO 




CM 














st 




CM 






CM 


i 






i 


ro 


z 






z 






z 






z 


I 




z 


1 




z 






ro 


o 






o 






o 






o 






o 






o 














































1— 






►— 






I— 






1- 












1— 








< 






< 






< 






< 






< 






< 








—1 


CM 




_l 


rs. 


in 


—1 


ro 






-o 


O 






ro 


_1 


rs. 


in 


st 


o 


i — 




o 


in 


CM 


o 


ro 


CO 


o 


CM 


CM 


o 


ro 


ro 


o 


o 


in 


so 


o 


ro 


CM 


CJ 


o 




CJ 


st 


CM 


CJ 


ro 


ro 


CJ 


o 


o 


CJ 




O 


CO 


oc 


o 


st 


oc 


IM 


r^ 


OC 


s3 




oc 

UJ 


ro 


|s- 


Q£ 
UJ 


in 


•o 


QC 
UJ 


st 


CM 




LU 
O- 


o 


in 


UJ 

CL 


rs. 


in 


UJ 
BL 


in 


O 


a. 




st 


a. 


st 


CM 


Q_ 


in 


LA 




















st 




CM 


1 






i 




i 


1 


£ 






















■ 
















CM 



c 



a 

o 
u 



Ol 

u 

H 
H 



st 

rs. 



rs. 
rs. 



CM 
CO 



is. 

CO 



st 
O 



o 

CM 



I— 


ro 


UJ 


i— 


ro 


UJ 


1— 


ro 


UJ 


1— 


ro 


UJ 


l— 


ro 


LU 




ro 


UJ 


AG 


< 






< 




CJ 


< 




u 






CJ 


< 




CJ 


< 




O 


cx 


<J 


'e 


< 


CJ 


'e 


< 


CJ 


'e 


< 


u 


'e 


< 


CJ 


'e 


< 


CJ 


E 


«t 


o 




> — ' 


OC 






oc 






cx: 




> — ' 


oc 




s> 


c* 






OC 


►— 


_J 




o 


_i 




o 


—I 




O 


—1 




o 


-J 




O 


_j 




o 


cn 


Q. 


UJ 


1— 


Cl 


UJ 


y— 


Q. 


UJ 


l— 


Q- 


UJ 


1— 


a. 


LU 


i— 


CL 


UJ 


1— 


3 


a. 


u 


oo 


CL 


CJ 


CO 


Q. 


CJ 


oo 


O- 


CJ 


CO 


CL 


CJ 


C/) 


CL 


CJ 


cn 




< 


RA 


3 


< 


RA 


3 


< 


RA 


3 


<t 


< 

oc 


3 


< 


RA 


3 


< 


RA 


3 


IN 


UJ 


o 


< 


UJ 


O 


< 


UJ 


o 


< 


UJ 


O 


< 


UJ 


o 


< 


UJ 


o 


< 




CJ 


i— 


1- 


CJ 


l— 


l— 


CJ 


1— 


1— 


CJ 


r- 


1- 


CJ 


»— 


l— 


CJ 


1— 


r- 


ac 


z 


oo 


_l 


z 


00 


_J 


z 


00 


—J 


z 


O) 


—1 


z 


V) 


I 


z 


10 




O 




3 


UJ 




3 


UJ 




3 


UJ 




3 


LU 




3 


UJ 




3 


LU 


oc 


oo 


< 


o 


oo 


< 


Q 


00 


< 


Q 


cn 


< 


O 


c/> 


< 


a 


c/> 


< 


a 


ER 


oo 


1— 


Q 


00 


►— 


O 


00 


l— 


o 


O) 


l— 


Q 


cn 


t— 


o 


CO 


t— 






>- 


— 1 


UJ 


>- 


_t 


UJ 


>- 


_l 


UJ 


>- 


—i 


LU 


>- 


—i 


UJ 


>- 


—l 


LU 


i— 


< 


UJ 


OC 


< 


LU 


OC 


< 


UJ 


oc 


< 


UJ 


OC 


< 


LU 


oc 


< 


UJ 


OC 


o 


o 


Q 


a. 


o 


O 


o_ 




o 


0l 




o 


CL 


o 


a 


a. 


a 


Q 


CL 


t— 



APPENDIX D 

SURFACE PLOTS OF ATRAZINE CONCENTRATION IN GROUNDWATER 



Date: 01/19/87 

08-09 




Figure D.l. Atrazine concentration in the groundwater on 1/19/87. 
Vertical bars indicate sampling locations. 



08-09 




Figure D.2. Atrazine concentration in the groundwater on 1/26/87. 
Vertical bars indicate sampling locations. 



201 



202 



Date: 02/02/87 




Figure D.3. Atrazine concentration in the groundwater on 2/02/87. 
Vertical bars indicate sampling locations. 



07_n Date: 02/09/87 

09-1 1 




Figure D.4. Atrazine concentration in the groundwater on 2/09/87. 
Vertical bars indicate sampling locations. 



203 




Figure D.5. Atrazine concentration in the groundwater on 2/16/87. 
Vertical bars indicate sampling locations. 




Figure D.6. Atrazine concentration in the groundwater on 2/23/87. 
Vertical bars indicate sampling locations. 



204 




O 



190 57 |sTA fjCE 



Figure D.7. Atrazine concentration in the groundwater on 3/02/87. 
Vertical bars indicate sampling locations. 



09-1 1 
08-09 



09-14 



Date: 03/09/87 




D| st ANC[ J5o 1 j^TW^. (m) 



Figure D.8. Atrazine concentration in the groundwater on 3/09/87. 
Vertical bars indicate sampling locations. 



205 




Figure D.9. Atrazine concentration in the groundwater on 3/16/87. 
Vertical bars indicate sampling locations. 




Figure D.10. Atrazine concentration in the groundwater on 3/23/87. 
Vertical bars indicate sampling locations. 



206 



Date: 03/31/87 




Figure D.ll. Atrazine concentration in the groundwater on 3/31/87. 
Vertical bars indicate sampling locations. 



09-1 1 




Figure D.12. Atrazine concentration in the groundwater on 4/06/87. 
Vertical bars indicate sampling locations. 



207 



10-13 




Figure D.13. Atrazine concentration in the groundwater on 4/13/87. 
Vertical bars indicate sampling locations. 



Date: 04/20/87 

08-10 




Figure D.14. Atrazine concentration in the groundwater on 4/20/87. 
Vertical bars indicate sampling locations. 



208 



Date: 04/30/87 




Figure D.15. Atrazine concentration in the groundwater on 4/30/87. 
Vertical bars indicate sampling locations. 



Date: 05/01/87 




Figure D.16. Atrazine concentration in the groundwater on 5/01/87. 
Vertical bars indicate sampling locations. 



209 




Figure D. 17. Atrazine concentration in the groundwater on 5/03/87. 
Vertical bars indicate sampling locations. 



08-09 




Figure D.18. Atrazine concentration in the groundwater on 5/05/87. 
Vertical bars indicate sampling locations. 



210 

08-09 




Figure D.19. Atrazine concentration in the groundwater on 5/08/87. 
Vertical bars indicate sampling locations. 




Figure D.20. Atrazine concentration in the groundwater on 5/13/87. 
Vertical bars indicate sampling locations. 



211 



10-09 




Figure D.21. Atrazine concentration in the groundwater on 5/18/87. 
Vertical bars indicate sampling locations. 



09-11 07-14 

Date: 05/25/87 




Figure D.22. Atrazine concentration in the groundwater on 5/25/87. 
Vertical bars indicate sampling locations. 



212 



09-1 1 




Figure D.23. Atrazine concentration in the groundwater on 6/01/87. 
Vertical bars indicate sampling locations. 



BIOGRAPHICAL SKETCH 
Matthew Clay Smith was born on September 27, 1957, in DeLand, 
Florida. He attended Deland Senior High School during 1973 and 1974. In 
1974 he attended Brevard College in Brevard, North Carolina, under the 
early admissions program. In 1977 he received the Associate of Science 
degree from Abraham Baldwin Agricultural College in Tifton, Georgia. In 
1980 he received the Bachelor of Science degree in agricultural 
engineering from the university of Georgia in Athens, Georgia. While 
attending the University of Georgia he participated in the cooperative 
education program by alternating quarters between academic courses and 
employment with the U. S. Environmental Protection Agency in Athens, 
Georgia. 

He attended graduate school at North Carolina State University in 
Raleigh, North Carolina, where he worked as a research assistant in the 
Department of Biological and Agricultural Engineering. He received the 
Master of Science degree in 1983. His thesis topic was a study of the 
water and energy use efficiency of drainage/sub irrigation systems. 

He worked as a research agricultural engineer in the Department of 
Agricultural Engineering, University of Georgia, Coastal Plain Experiment 
Station in Tifton, Georgia, from 1982 to 1984. 

He attended the University of Florida in Gainesville, Florida, where 
he worked as a research assistant in the Agricultural Engineering 
Department from May, 1984, through September, 1987. 



213 



214 

In October, 1987, he began employment with the University of 
Georgia, Coastal Plain Experiment Station in Tifton, Georgia, where he 
holds the title of Assistant Professor in the Department of Agricultural 
Engineering. 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 




A. B. (Del) Bottcher, Chairman 
Professor of Agricultural Engineering 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 

Kenneth L. Campbell, Codhairman 
Associate Professor of Agricultural 
Engineering 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. — \ i 

i /?uw „ C Mi.U<k. 

Wayne <z. Huber 

Professor of Environmental Engineering 
Sciences 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 



Doctor of Philosophy. 




Professor of Soil Science 



I certify that I have read this study and that in my opinion it 
conforms to acceptable standards of scholarly presentation and is fully 
adequate, in scope and quality, as a dissertation for the degree of 
Doctor of Philosophy. 

E. Dale Hireadgill / 
Professor and Division Chairman 

Agricultural Engineering, University 

of Georgia 



This dissertation was submitted to the Graduate Faculty of the 
College of Engineering and to the Graduate School and was accepted as 
partial fulfillment of the requirements for the degree of Doctor of 
Philosophy. 



December, 1988 




Dean, Graduate School